Commit
•
1e99df2
1
Parent(s):
274419a
Evaluation on the test set completed on 2024_09_03.
Browse files- README.md +208 -0
- all_results.json +17 -0
- logs/events.out.tfevents.1725102544.datavisu2 +2 -2
- logs/events.out.tfevents.1725352882.datavisu2 +3 -0
- model.safetensors +1 -1
- test_results.json +12 -0
- train_results.json +9 -0
- trainer_state.json +2470 -0
README.md
ADDED
|
@@ -0,0 +1,208 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: facebook/dinov2-small
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
metrics:
|
| 7 |
+
- accuracy
|
| 8 |
+
model-index:
|
| 9 |
+
- name: DinoVdeau-small-2024_08_31-batch-size32_epochs150_freeze
|
| 10 |
+
results: []
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 14 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 15 |
+
|
| 16 |
+
# DinoVdeau-small-2024_08_31-batch-size32_epochs150_freeze
|
| 17 |
+
|
| 18 |
+
This model is a fine-tuned version of [facebook/dinov2-small](https://huggingface.co/facebook/dinov2-small) on the None dataset.
|
| 19 |
+
It achieves the following results on the evaluation set:
|
| 20 |
+
- Loss: 0.1320
|
| 21 |
+
- F1 Micro: 0.8009
|
| 22 |
+
- F1 Macro: 0.6614
|
| 23 |
+
- Roc Auc: 0.8649
|
| 24 |
+
- Accuracy: 0.2903
|
| 25 |
+
- Learning Rate: 0.0000
|
| 26 |
+
|
| 27 |
+
## Model description
|
| 28 |
+
|
| 29 |
+
More information needed
|
| 30 |
+
|
| 31 |
+
## Intended uses & limitations
|
| 32 |
+
|
| 33 |
+
More information needed
|
| 34 |
+
|
| 35 |
+
## Training and evaluation data
|
| 36 |
+
|
| 37 |
+
More information needed
|
| 38 |
+
|
| 39 |
+
## Training procedure
|
| 40 |
+
|
| 41 |
+
### Training hyperparameters
|
| 42 |
+
|
| 43 |
+
The following hyperparameters were used during training:
|
| 44 |
+
- learning_rate: 0.001
|
| 45 |
+
- train_batch_size: 32
|
| 46 |
+
- eval_batch_size: 32
|
| 47 |
+
- seed: 42
|
| 48 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 49 |
+
- lr_scheduler_type: linear
|
| 50 |
+
- num_epochs: 150
|
| 51 |
+
- mixed_precision_training: Native AMP
|
| 52 |
+
|
| 53 |
+
### Training results
|
| 54 |
+
|
| 55 |
+
| Training Loss | Epoch | Step | Validation Loss | F1 Micro | F1 Macro | Roc Auc | Accuracy | Rate |
|
| 56 |
+
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:--------:|:-------:|:--------:|:------:|
|
| 57 |
+
| No log | 1.0 | 273 | 0.1957 | 0.7089 | 0.4059 | 0.8061 | 0.1906 | 0.001 |
|
| 58 |
+
| 0.3189 | 2.0 | 546 | 0.1720 | 0.7381 | 0.4868 | 0.8255 | 0.2193 | 0.001 |
|
| 59 |
+
| 0.3189 | 3.0 | 819 | 0.1621 | 0.7579 | 0.5587 | 0.8388 | 0.2322 | 0.001 |
|
| 60 |
+
| 0.1897 | 4.0 | 1092 | 0.1595 | 0.7463 | 0.5562 | 0.8221 | 0.2249 | 0.001 |
|
| 61 |
+
| 0.1897 | 5.0 | 1365 | 0.1569 | 0.7511 | 0.5723 | 0.8245 | 0.2315 | 0.001 |
|
| 62 |
+
| 0.1808 | 6.0 | 1638 | 0.1530 | 0.7635 | 0.5787 | 0.8365 | 0.2363 | 0.001 |
|
| 63 |
+
| 0.1808 | 7.0 | 1911 | 0.1523 | 0.7652 | 0.5982 | 0.8389 | 0.2335 | 0.001 |
|
| 64 |
+
| 0.1763 | 8.0 | 2184 | 0.1531 | 0.7655 | 0.5880 | 0.8377 | 0.2419 | 0.001 |
|
| 65 |
+
| 0.1763 | 9.0 | 2457 | 0.1499 | 0.7700 | 0.6069 | 0.8431 | 0.2401 | 0.001 |
|
| 66 |
+
| 0.1735 | 10.0 | 2730 | 0.1510 | 0.7606 | 0.5829 | 0.8277 | 0.2439 | 0.001 |
|
| 67 |
+
| 0.1723 | 11.0 | 3003 | 0.1521 | 0.7690 | 0.5976 | 0.8400 | 0.2505 | 0.001 |
|
| 68 |
+
| 0.1723 | 12.0 | 3276 | 0.1503 | 0.7760 | 0.6074 | 0.8527 | 0.2443 | 0.001 |
|
| 69 |
+
| 0.1719 | 13.0 | 3549 | 0.1504 | 0.7624 | 0.6003 | 0.8302 | 0.2439 | 0.001 |
|
| 70 |
+
| 0.1719 | 14.0 | 3822 | 0.1497 | 0.7644 | 0.6028 | 0.8343 | 0.2446 | 0.001 |
|
| 71 |
+
| 0.1702 | 15.0 | 4095 | 0.1475 | 0.7752 | 0.6066 | 0.8446 | 0.2512 | 0.001 |
|
| 72 |
+
| 0.1702 | 16.0 | 4368 | 0.1500 | 0.7646 | 0.5838 | 0.8321 | 0.2464 | 0.001 |
|
| 73 |
+
| 0.1696 | 17.0 | 4641 | 0.1530 | 0.7720 | 0.6073 | 0.8464 | 0.2457 | 0.001 |
|
| 74 |
+
| 0.1696 | 18.0 | 4914 | 0.1491 | 0.7752 | 0.6143 | 0.8475 | 0.2439 | 0.001 |
|
| 75 |
+
| 0.1717 | 19.0 | 5187 | 0.1495 | 0.7740 | 0.6075 | 0.8484 | 0.2346 | 0.001 |
|
| 76 |
+
| 0.1717 | 20.0 | 5460 | 0.1487 | 0.7637 | 0.5956 | 0.8322 | 0.2453 | 0.001 |
|
| 77 |
+
| 0.1705 | 21.0 | 5733 | 0.1471 | 0.7805 | 0.6165 | 0.8540 | 0.2474 | 0.001 |
|
| 78 |
+
| 0.1706 | 22.0 | 6006 | 0.1509 | 0.7754 | 0.6074 | 0.8494 | 0.2453 | 0.001 |
|
| 79 |
+
| 0.1706 | 23.0 | 6279 | 0.1502 | 0.7719 | 0.6127 | 0.8388 | 0.2429 | 0.001 |
|
| 80 |
+
| 0.1699 | 24.0 | 6552 | 0.1497 | 0.7699 | 0.5849 | 0.8406 | 0.2401 | 0.001 |
|
| 81 |
+
| 0.1699 | 25.0 | 6825 | 0.1470 | 0.7761 | 0.6035 | 0.8459 | 0.2426 | 0.001 |
|
| 82 |
+
| 0.1694 | 26.0 | 7098 | 0.1481 | 0.7751 | 0.6065 | 0.8466 | 0.2422 | 0.001 |
|
| 83 |
+
| 0.1694 | 27.0 | 7371 | 0.1458 | 0.7689 | 0.6136 | 0.8357 | 0.2474 | 0.001 |
|
| 84 |
+
| 0.17 | 28.0 | 7644 | 0.1454 | 0.7751 | 0.6077 | 0.8441 | 0.2446 | 0.001 |
|
| 85 |
+
| 0.17 | 29.0 | 7917 | 0.1494 | 0.7735 | 0.6108 | 0.8491 | 0.2457 | 0.001 |
|
| 86 |
+
| 0.1685 | 30.0 | 8190 | 0.1455 | 0.7705 | 0.5983 | 0.8366 | 0.2498 | 0.001 |
|
| 87 |
+
| 0.1685 | 31.0 | 8463 | 0.1454 | 0.7785 | 0.6069 | 0.8495 | 0.2533 | 0.001 |
|
| 88 |
+
| 0.1687 | 32.0 | 8736 | 0.1466 | 0.7746 | 0.6145 | 0.8461 | 0.2453 | 0.001 |
|
| 89 |
+
| 0.1679 | 33.0 | 9009 | 0.1446 | 0.7770 | 0.6125 | 0.8439 | 0.2540 | 0.001 |
|
| 90 |
+
| 0.1679 | 34.0 | 9282 | 0.1468 | 0.7781 | 0.6168 | 0.8470 | 0.2446 | 0.001 |
|
| 91 |
+
| 0.168 | 35.0 | 9555 | 0.1486 | 0.7767 | 0.6193 | 0.8452 | 0.2495 | 0.001 |
|
| 92 |
+
| 0.168 | 36.0 | 9828 | 0.1464 | 0.7719 | 0.6093 | 0.8391 | 0.2488 | 0.001 |
|
| 93 |
+
| 0.169 | 37.0 | 10101 | 0.1448 | 0.7734 | 0.6127 | 0.8402 | 0.2498 | 0.001 |
|
| 94 |
+
| 0.169 | 38.0 | 10374 | 0.1451 | 0.7815 | 0.6110 | 0.8526 | 0.2523 | 0.001 |
|
| 95 |
+
| 0.167 | 39.0 | 10647 | 0.1447 | 0.7824 | 0.6272 | 0.8563 | 0.2498 | 0.001 |
|
| 96 |
+
| 0.167 | 40.0 | 10920 | 0.1482 | 0.7837 | 0.6266 | 0.8537 | 0.2536 | 0.0001 |
|
| 97 |
+
| 0.1652 | 41.0 | 11193 | 0.1414 | 0.7833 | 0.6324 | 0.8483 | 0.2616 | 0.0001 |
|
| 98 |
+
| 0.1652 | 42.0 | 11466 | 0.1398 | 0.7884 | 0.6372 | 0.8546 | 0.2620 | 0.0001 |
|
| 99 |
+
| 0.1608 | 43.0 | 11739 | 0.1411 | 0.7871 | 0.6367 | 0.8537 | 0.2640 | 0.0001 |
|
| 100 |
+
| 0.1596 | 44.0 | 12012 | 0.1390 | 0.7879 | 0.6257 | 0.8537 | 0.2613 | 0.0001 |
|
| 101 |
+
| 0.1596 | 45.0 | 12285 | 0.1386 | 0.7894 | 0.6421 | 0.8539 | 0.2665 | 0.0001 |
|
| 102 |
+
| 0.1582 | 46.0 | 12558 | 0.1396 | 0.7874 | 0.6283 | 0.8522 | 0.2665 | 0.0001 |
|
| 103 |
+
| 0.1582 | 47.0 | 12831 | 0.1387 | 0.7864 | 0.6287 | 0.8500 | 0.2637 | 0.0001 |
|
| 104 |
+
| 0.1584 | 48.0 | 13104 | 0.1378 | 0.7913 | 0.6335 | 0.8572 | 0.2678 | 0.0001 |
|
| 105 |
+
| 0.1584 | 49.0 | 13377 | 0.1377 | 0.7934 | 0.6382 | 0.8603 | 0.2640 | 0.0001 |
|
| 106 |
+
| 0.157 | 50.0 | 13650 | 0.1376 | 0.7918 | 0.6363 | 0.8570 | 0.2675 | 0.0001 |
|
| 107 |
+
| 0.157 | 51.0 | 13923 | 0.1375 | 0.7929 | 0.6427 | 0.8597 | 0.2661 | 0.0001 |
|
| 108 |
+
| 0.1567 | 52.0 | 14196 | 0.1377 | 0.7871 | 0.6368 | 0.8507 | 0.2658 | 0.0001 |
|
| 109 |
+
| 0.1567 | 53.0 | 14469 | 0.1374 | 0.7929 | 0.6406 | 0.8601 | 0.2692 | 0.0001 |
|
| 110 |
+
| 0.1571 | 54.0 | 14742 | 0.1369 | 0.7921 | 0.6412 | 0.8562 | 0.2717 | 0.0001 |
|
| 111 |
+
| 0.1548 | 55.0 | 15015 | 0.1370 | 0.7914 | 0.6378 | 0.8558 | 0.2703 | 0.0001 |
|
| 112 |
+
| 0.1548 | 56.0 | 15288 | 0.1365 | 0.7931 | 0.6425 | 0.8602 | 0.2644 | 0.0001 |
|
| 113 |
+
| 0.155 | 57.0 | 15561 | 0.1368 | 0.7926 | 0.6382 | 0.8588 | 0.2675 | 0.0001 |
|
| 114 |
+
| 0.155 | 58.0 | 15834 | 0.1365 | 0.7916 | 0.6374 | 0.8553 | 0.2675 | 0.0001 |
|
| 115 |
+
| 0.155 | 59.0 | 16107 | 0.1364 | 0.7922 | 0.6429 | 0.8565 | 0.2675 | 0.0001 |
|
| 116 |
+
| 0.155 | 60.0 | 16380 | 0.1369 | 0.7883 | 0.6358 | 0.8515 | 0.2651 | 0.0001 |
|
| 117 |
+
| 0.1546 | 61.0 | 16653 | 0.1364 | 0.7946 | 0.6504 | 0.8589 | 0.2713 | 0.0001 |
|
| 118 |
+
| 0.1546 | 62.0 | 16926 | 0.1356 | 0.7932 | 0.6442 | 0.8575 | 0.2751 | 0.0001 |
|
| 119 |
+
| 0.1536 | 63.0 | 17199 | 0.1355 | 0.7966 | 0.6516 | 0.8611 | 0.2737 | 0.0001 |
|
| 120 |
+
| 0.1536 | 64.0 | 17472 | 0.1359 | 0.7934 | 0.6450 | 0.8578 | 0.2678 | 0.0001 |
|
| 121 |
+
| 0.1544 | 65.0 | 17745 | 0.1357 | 0.7936 | 0.6455 | 0.8572 | 0.2706 | 0.0001 |
|
| 122 |
+
| 0.1529 | 66.0 | 18018 | 0.1357 | 0.7946 | 0.6477 | 0.8595 | 0.2713 | 0.0001 |
|
| 123 |
+
| 0.1529 | 67.0 | 18291 | 0.1353 | 0.7966 | 0.6544 | 0.8623 | 0.2755 | 0.0001 |
|
| 124 |
+
| 0.1528 | 68.0 | 18564 | 0.1353 | 0.7956 | 0.6519 | 0.8608 | 0.2734 | 0.0001 |
|
| 125 |
+
| 0.1528 | 69.0 | 18837 | 0.1347 | 0.7966 | 0.6516 | 0.8603 | 0.2699 | 0.0001 |
|
| 126 |
+
| 0.1528 | 70.0 | 19110 | 0.1350 | 0.7945 | 0.6442 | 0.8575 | 0.2720 | 0.0001 |
|
| 127 |
+
| 0.1528 | 71.0 | 19383 | 0.1350 | 0.7933 | 0.6442 | 0.8557 | 0.2723 | 0.0001 |
|
| 128 |
+
| 0.1522 | 72.0 | 19656 | 0.1345 | 0.7970 | 0.6485 | 0.8605 | 0.2758 | 0.0001 |
|
| 129 |
+
| 0.1522 | 73.0 | 19929 | 0.1342 | 0.7977 | 0.6519 | 0.8616 | 0.2762 | 0.0001 |
|
| 130 |
+
| 0.1523 | 74.0 | 20202 | 0.1350 | 0.7915 | 0.6413 | 0.8520 | 0.2751 | 0.0001 |
|
| 131 |
+
| 0.1523 | 75.0 | 20475 | 0.1346 | 0.7947 | 0.6485 | 0.8572 | 0.2751 | 0.0001 |
|
| 132 |
+
| 0.1521 | 76.0 | 20748 | 0.1344 | 0.7965 | 0.6478 | 0.8598 | 0.2758 | 0.0001 |
|
| 133 |
+
| 0.1515 | 77.0 | 21021 | 0.1346 | 0.7978 | 0.6537 | 0.8623 | 0.2775 | 0.0001 |
|
| 134 |
+
| 0.1515 | 78.0 | 21294 | 0.1341 | 0.7978 | 0.6543 | 0.8635 | 0.2775 | 0.0001 |
|
| 135 |
+
| 0.1514 | 79.0 | 21567 | 0.1340 | 0.7953 | 0.6523 | 0.8574 | 0.2741 | 0.0001 |
|
| 136 |
+
| 0.1514 | 80.0 | 21840 | 0.1344 | 0.7993 | 0.6546 | 0.8653 | 0.2782 | 0.0001 |
|
| 137 |
+
| 0.1516 | 81.0 | 22113 | 0.1341 | 0.7967 | 0.6560 | 0.8576 | 0.2758 | 0.0001 |
|
| 138 |
+
| 0.1516 | 82.0 | 22386 | 0.1341 | 0.7948 | 0.6454 | 0.8555 | 0.2765 | 0.0001 |
|
| 139 |
+
| 0.149 | 83.0 | 22659 | 0.1351 | 0.7924 | 0.6460 | 0.8543 | 0.2703 | 0.0001 |
|
| 140 |
+
| 0.149 | 84.0 | 22932 | 0.1339 | 0.7957 | 0.6512 | 0.8586 | 0.2755 | 0.0001 |
|
| 141 |
+
| 0.1515 | 85.0 | 23205 | 0.1334 | 0.7991 | 0.6532 | 0.8620 | 0.2793 | 0.0001 |
|
| 142 |
+
| 0.1515 | 86.0 | 23478 | 0.1334 | 0.7988 | 0.6596 | 0.8625 | 0.2748 | 0.0001 |
|
| 143 |
+
| 0.1495 | 87.0 | 23751 | 0.1340 | 0.7956 | 0.6467 | 0.8591 | 0.2744 | 0.0001 |
|
| 144 |
+
| 0.1496 | 88.0 | 24024 | 0.1336 | 0.7982 | 0.6483 | 0.8620 | 0.2748 | 0.0001 |
|
| 145 |
+
| 0.1496 | 89.0 | 24297 | 0.1337 | 0.8015 | 0.6585 | 0.8672 | 0.2807 | 0.0001 |
|
| 146 |
+
| 0.1493 | 90.0 | 24570 | 0.1333 | 0.8011 | 0.6621 | 0.8661 | 0.2772 | 0.0001 |
|
| 147 |
+
| 0.1493 | 91.0 | 24843 | 0.1337 | 0.7957 | 0.6529 | 0.8563 | 0.2782 | 0.0001 |
|
| 148 |
+
| 0.1496 | 92.0 | 25116 | 0.1335 | 0.7961 | 0.6514 | 0.8574 | 0.2755 | 0.0001 |
|
| 149 |
+
| 0.1496 | 93.0 | 25389 | 0.1331 | 0.8002 | 0.6560 | 0.8648 | 0.2758 | 0.0001 |
|
| 150 |
+
| 0.1493 | 94.0 | 25662 | 0.1333 | 0.7995 | 0.6554 | 0.8643 | 0.2758 | 0.0001 |
|
| 151 |
+
| 0.1493 | 95.0 | 25935 | 0.1331 | 0.7980 | 0.6580 | 0.8606 | 0.2758 | 0.0001 |
|
| 152 |
+
| 0.1482 | 96.0 | 26208 | 0.1328 | 0.7993 | 0.6556 | 0.8631 | 0.2751 | 0.0001 |
|
| 153 |
+
| 0.1482 | 97.0 | 26481 | 0.1333 | 0.7977 | 0.6493 | 0.8589 | 0.2782 | 0.0001 |
|
| 154 |
+
| 0.1497 | 98.0 | 26754 | 0.1327 | 0.7996 | 0.6600 | 0.8647 | 0.2755 | 0.0001 |
|
| 155 |
+
| 0.1489 | 99.0 | 27027 | 0.1325 | 0.7979 | 0.6590 | 0.8608 | 0.2717 | 0.0001 |
|
| 156 |
+
| 0.1489 | 100.0 | 27300 | 0.1329 | 0.7971 | 0.6570 | 0.8585 | 0.2762 | 0.0001 |
|
| 157 |
+
| 0.1482 | 101.0 | 27573 | 0.1327 | 0.7992 | 0.6580 | 0.8611 | 0.2821 | 0.0001 |
|
| 158 |
+
| 0.1482 | 102.0 | 27846 | 0.1326 | 0.7987 | 0.6543 | 0.8608 | 0.2817 | 0.0001 |
|
| 159 |
+
| 0.1474 | 103.0 | 28119 | 0.1325 | 0.7994 | 0.6518 | 0.8621 | 0.2803 | 0.0001 |
|
| 160 |
+
| 0.1474 | 104.0 | 28392 | 0.1332 | 0.8011 | 0.6613 | 0.8647 | 0.2775 | 0.0001 |
|
| 161 |
+
| 0.1472 | 105.0 | 28665 | 0.1322 | 0.8013 | 0.6636 | 0.8652 | 0.2831 | 0.0001 |
|
| 162 |
+
| 0.1472 | 106.0 | 28938 | 0.1324 | 0.8010 | 0.6588 | 0.8633 | 0.2831 | 0.0001 |
|
| 163 |
+
| 0.148 | 107.0 | 29211 | 0.1336 | 0.7986 | 0.6506 | 0.8619 | 0.2786 | 0.0001 |
|
| 164 |
+
| 0.148 | 108.0 | 29484 | 0.1327 | 0.7996 | 0.6501 | 0.8615 | 0.2796 | 0.0001 |
|
| 165 |
+
| 0.1477 | 109.0 | 29757 | 0.1318 | 0.8000 | 0.6580 | 0.8613 | 0.2807 | 0.0001 |
|
| 166 |
+
| 0.1479 | 110.0 | 30030 | 0.1326 | 0.7997 | 0.6582 | 0.8626 | 0.2803 | 0.0001 |
|
| 167 |
+
| 0.1479 | 111.0 | 30303 | 0.1319 | 0.8013 | 0.6609 | 0.8638 | 0.2786 | 0.0001 |
|
| 168 |
+
| 0.1466 | 112.0 | 30576 | 0.1322 | 0.8019 | 0.6595 | 0.8659 | 0.2810 | 0.0001 |
|
| 169 |
+
| 0.1466 | 113.0 | 30849 | 0.1321 | 0.8025 | 0.6592 | 0.8667 | 0.2800 | 0.0001 |
|
| 170 |
+
| 0.1474 | 114.0 | 31122 | 0.1320 | 0.8025 | 0.6631 | 0.8662 | 0.2824 | 0.0001 |
|
| 171 |
+
| 0.1474 | 115.0 | 31395 | 0.1319 | 0.8004 | 0.6598 | 0.8625 | 0.2838 | 0.0001 |
|
| 172 |
+
| 0.1468 | 116.0 | 31668 | 0.1319 | 0.8022 | 0.6627 | 0.8643 | 0.2845 | 1e-05 |
|
| 173 |
+
| 0.1468 | 117.0 | 31941 | 0.1318 | 0.8013 | 0.6604 | 0.8634 | 0.2821 | 1e-05 |
|
| 174 |
+
| 0.1455 | 118.0 | 32214 | 0.1316 | 0.8002 | 0.6590 | 0.8616 | 0.2796 | 1e-05 |
|
| 175 |
+
| 0.1455 | 119.0 | 32487 | 0.1319 | 0.8037 | 0.6608 | 0.8678 | 0.2827 | 1e-05 |
|
| 176 |
+
| 0.1451 | 120.0 | 32760 | 0.1316 | 0.8036 | 0.6615 | 0.8662 | 0.2814 | 1e-05 |
|
| 177 |
+
| 0.1454 | 121.0 | 33033 | 0.1318 | 0.8013 | 0.6611 | 0.8635 | 0.2810 | 1e-05 |
|
| 178 |
+
| 0.1454 | 122.0 | 33306 | 0.1322 | 0.8050 | 0.6647 | 0.8692 | 0.2817 | 1e-05 |
|
| 179 |
+
| 0.145 | 123.0 | 33579 | 0.1319 | 0.8010 | 0.6605 | 0.8618 | 0.2817 | 1e-05 |
|
| 180 |
+
| 0.145 | 124.0 | 33852 | 0.1314 | 0.8019 | 0.6622 | 0.8638 | 0.2807 | 1e-05 |
|
| 181 |
+
| 0.1459 | 125.0 | 34125 | 0.1314 | 0.8043 | 0.6641 | 0.8672 | 0.2862 | 1e-05 |
|
| 182 |
+
| 0.1459 | 126.0 | 34398 | 0.1310 | 0.8042 | 0.6630 | 0.8670 | 0.2862 | 1e-05 |
|
| 183 |
+
| 0.1439 | 127.0 | 34671 | 0.1315 | 0.8038 | 0.6598 | 0.8673 | 0.2859 | 1e-05 |
|
| 184 |
+
| 0.1439 | 128.0 | 34944 | 0.1311 | 0.8042 | 0.6682 | 0.8674 | 0.2869 | 1e-05 |
|
| 185 |
+
| 0.1446 | 129.0 | 35217 | 0.1310 | 0.8035 | 0.6653 | 0.8665 | 0.2827 | 1e-05 |
|
| 186 |
+
| 0.1446 | 130.0 | 35490 | 0.1310 | 0.8034 | 0.6657 | 0.8668 | 0.2866 | 1e-05 |
|
| 187 |
+
| 0.1449 | 131.0 | 35763 | 0.1313 | 0.8052 | 0.6709 | 0.8699 | 0.2834 | 1e-05 |
|
| 188 |
+
| 0.1442 | 132.0 | 36036 | 0.1315 | 0.7986 | 0.6558 | 0.8595 | 0.2807 | 1e-05 |
|
| 189 |
+
| 0.1442 | 133.0 | 36309 | 0.1311 | 0.8052 | 0.6689 | 0.8692 | 0.2879 | 1e-05 |
|
| 190 |
+
| 0.1443 | 134.0 | 36582 | 0.1309 | 0.8021 | 0.6648 | 0.8640 | 0.2827 | 1e-05 |
|
| 191 |
+
| 0.1443 | 135.0 | 36855 | 0.1315 | 0.8038 | 0.6684 | 0.8665 | 0.2869 | 1e-05 |
|
| 192 |
+
| 0.1438 | 136.0 | 37128 | 0.1315 | 0.8025 | 0.6590 | 0.8634 | 0.2827 | 1e-05 |
|
| 193 |
+
| 0.1438 | 137.0 | 37401 | 0.1311 | 0.8036 | 0.6667 | 0.8648 | 0.2859 | 1e-05 |
|
| 194 |
+
| 0.1452 | 138.0 | 37674 | 0.1312 | 0.8035 | 0.6666 | 0.8661 | 0.2845 | 1e-05 |
|
| 195 |
+
| 0.1452 | 139.0 | 37947 | 0.1310 | 0.8053 | 0.6661 | 0.8689 | 0.2897 | 1e-05 |
|
| 196 |
+
| 0.144 | 140.0 | 38220 | 0.1317 | 0.8020 | 0.6635 | 0.8643 | 0.2834 | 1e-05 |
|
| 197 |
+
| 0.144 | 141.0 | 38493 | 0.1309 | 0.8047 | 0.6688 | 0.8673 | 0.2876 | 0.0000 |
|
| 198 |
+
| 0.1445 | 142.0 | 38766 | 0.1310 | 0.8042 | 0.6643 | 0.8657 | 0.2859 | 0.0000 |
|
| 199 |
+
| 0.1441 | 143.0 | 39039 | 0.1314 | 0.8019 | 0.6623 | 0.8635 | 0.2872 | 0.0000 |
|
| 200 |
+
| 0.1441 | 144.0 | 39312 | 0.1312 | 0.8025 | 0.6648 | 0.8649 | 0.2838 | 0.0000 |
|
| 201 |
+
|
| 202 |
+
|
| 203 |
+
### Framework versions
|
| 204 |
+
|
| 205 |
+
- Transformers 4.41.1
|
| 206 |
+
- Pytorch 2.3.0+cu121
|
| 207 |
+
- Datasets 2.19.1
|
| 208 |
+
- Tokenizers 0.19.1
|
all_results.json
ADDED
|
@@ -0,0 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 144.0,
|
| 3 |
+
"eval_accuracy": 0.2903114186851211,
|
| 4 |
+
"eval_f1_macro": 0.661445626763805,
|
| 5 |
+
"eval_f1_micro": 0.800865984633018,
|
| 6 |
+
"eval_loss": 0.13202470541000366,
|
| 7 |
+
"eval_roc_auc": 0.8648958366235343,
|
| 8 |
+
"eval_runtime": 442.1525,
|
| 9 |
+
"eval_samples_per_second": 6.536,
|
| 10 |
+
"eval_steps_per_second": 0.206,
|
| 11 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 12 |
+
"total_flos": 1.3598709030716368e+20,
|
| 13 |
+
"train_loss": 0.157796386979584,
|
| 14 |
+
"train_runtime": 249885.5342,
|
| 15 |
+
"train_samples_per_second": 5.232,
|
| 16 |
+
"train_steps_per_second": 0.164
|
| 17 |
+
}
|
logs/events.out.tfevents.1725102544.datavisu2
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c78f8d588e03c07136f112bd0733c82be709d721e766adc23738234bdd650852
|
| 3 |
+
size 102137
|
logs/events.out.tfevents.1725352882.datavisu2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:02294ba166290ffcd374d6f7b2c0739495a8f0636b72b1daedb5427d8a3377fc
|
| 3 |
+
size 40
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 89762948
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab1f92930801f6ecde589d4e84c8913d6ecae132a9ee667fe3fdfbe078bdafb3
|
| 3 |
size 89762948
|
test_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 144.0,
|
| 3 |
+
"eval_accuracy": 0.2903114186851211,
|
| 4 |
+
"eval_f1_macro": 0.661445626763805,
|
| 5 |
+
"eval_f1_micro": 0.800865984633018,
|
| 6 |
+
"eval_loss": 0.13202470541000366,
|
| 7 |
+
"eval_roc_auc": 0.8648958366235343,
|
| 8 |
+
"eval_runtime": 442.1525,
|
| 9 |
+
"eval_samples_per_second": 6.536,
|
| 10 |
+
"eval_steps_per_second": 0.206,
|
| 11 |
+
"learning_rate": 1.0000000000000002e-06
|
| 12 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 144.0,
|
| 3 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 4 |
+
"total_flos": 1.3598709030716368e+20,
|
| 5 |
+
"train_loss": 0.157796386979584,
|
| 6 |
+
"train_runtime": 249885.5342,
|
| 7 |
+
"train_samples_per_second": 5.232,
|
| 8 |
+
"train_steps_per_second": 0.164
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2470 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.1308571696281433,
|
| 3 |
+
"best_model_checkpoint": "/home/datawork-iot-nos/Seatizen/models/multilabel/fine_scale/DinoVdeau-small-2024_08_31-batch-size32_epochs150_freeze/checkpoint-36582",
|
| 4 |
+
"epoch": 144.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 39312,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"eval_accuracy": 0.19057519057519057,
|
| 14 |
+
"eval_f1_macro": 0.4058921954514261,
|
| 15 |
+
"eval_f1_micro": 0.7088941673264713,
|
| 16 |
+
"eval_loss": 0.19568666815757751,
|
| 17 |
+
"eval_roc_auc": 0.8060676064167129,
|
| 18 |
+
"eval_runtime": 426.0483,
|
| 19 |
+
"eval_samples_per_second": 6.774,
|
| 20 |
+
"eval_steps_per_second": 0.214,
|
| 21 |
+
"learning_rate": 0.001,
|
| 22 |
+
"step": 273
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 1.8315018315018317,
|
| 26 |
+
"grad_norm": 0.30737248063087463,
|
| 27 |
+
"learning_rate": 0.001,
|
| 28 |
+
"loss": 0.3189,
|
| 29 |
+
"step": 500
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 2.0,
|
| 33 |
+
"eval_accuracy": 0.21933471933471935,
|
| 34 |
+
"eval_f1_macro": 0.4867943512801917,
|
| 35 |
+
"eval_f1_micro": 0.738139514768845,
|
| 36 |
+
"eval_loss": 0.17198018729686737,
|
| 37 |
+
"eval_roc_auc": 0.8255075095586444,
|
| 38 |
+
"eval_runtime": 425.0166,
|
| 39 |
+
"eval_samples_per_second": 6.79,
|
| 40 |
+
"eval_steps_per_second": 0.214,
|
| 41 |
+
"learning_rate": 0.001,
|
| 42 |
+
"step": 546
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 3.0,
|
| 46 |
+
"eval_accuracy": 0.23215523215523215,
|
| 47 |
+
"eval_f1_macro": 0.5587016500092944,
|
| 48 |
+
"eval_f1_micro": 0.7578947368421052,
|
| 49 |
+
"eval_loss": 0.16209888458251953,
|
| 50 |
+
"eval_roc_auc": 0.8387630797560628,
|
| 51 |
+
"eval_runtime": 425.9119,
|
| 52 |
+
"eval_samples_per_second": 6.776,
|
| 53 |
+
"eval_steps_per_second": 0.214,
|
| 54 |
+
"learning_rate": 0.001,
|
| 55 |
+
"step": 819
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 3.663003663003663,
|
| 59 |
+
"grad_norm": 0.2619726359844208,
|
| 60 |
+
"learning_rate": 0.001,
|
| 61 |
+
"loss": 0.1897,
|
| 62 |
+
"step": 1000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 4.0,
|
| 66 |
+
"eval_accuracy": 0.22487872487872487,
|
| 67 |
+
"eval_f1_macro": 0.5561953540051209,
|
| 68 |
+
"eval_f1_micro": 0.7463059684835497,
|
| 69 |
+
"eval_loss": 0.15948981046676636,
|
| 70 |
+
"eval_roc_auc": 0.8221271753092407,
|
| 71 |
+
"eval_runtime": 423.9484,
|
| 72 |
+
"eval_samples_per_second": 6.807,
|
| 73 |
+
"eval_steps_per_second": 0.215,
|
| 74 |
+
"learning_rate": 0.001,
|
| 75 |
+
"step": 1092
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 5.0,
|
| 79 |
+
"eval_accuracy": 0.23146223146223147,
|
| 80 |
+
"eval_f1_macro": 0.5723046956548954,
|
| 81 |
+
"eval_f1_micro": 0.7510718113612004,
|
| 82 |
+
"eval_loss": 0.15691693127155304,
|
| 83 |
+
"eval_roc_auc": 0.8244935635420478,
|
| 84 |
+
"eval_runtime": 423.6041,
|
| 85 |
+
"eval_samples_per_second": 6.813,
|
| 86 |
+
"eval_steps_per_second": 0.215,
|
| 87 |
+
"learning_rate": 0.001,
|
| 88 |
+
"step": 1365
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 5.4945054945054945,
|
| 92 |
+
"grad_norm": 0.17114631831645966,
|
| 93 |
+
"learning_rate": 0.001,
|
| 94 |
+
"loss": 0.1808,
|
| 95 |
+
"step": 1500
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 6.0,
|
| 99 |
+
"eval_accuracy": 0.2363132363132363,
|
| 100 |
+
"eval_f1_macro": 0.5786669115862841,
|
| 101 |
+
"eval_f1_micro": 0.7634727923836142,
|
| 102 |
+
"eval_loss": 0.15302371978759766,
|
| 103 |
+
"eval_roc_auc": 0.8365257318814997,
|
| 104 |
+
"eval_runtime": 427.5566,
|
| 105 |
+
"eval_samples_per_second": 6.75,
|
| 106 |
+
"eval_steps_per_second": 0.213,
|
| 107 |
+
"learning_rate": 0.001,
|
| 108 |
+
"step": 1638
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 7.0,
|
| 112 |
+
"eval_accuracy": 0.23354123354123354,
|
| 113 |
+
"eval_f1_macro": 0.5981729145672101,
|
| 114 |
+
"eval_f1_micro": 0.7651630269613162,
|
| 115 |
+
"eval_loss": 0.1523299366235733,
|
| 116 |
+
"eval_roc_auc": 0.838924594824006,
|
| 117 |
+
"eval_runtime": 430.1478,
|
| 118 |
+
"eval_samples_per_second": 6.709,
|
| 119 |
+
"eval_steps_per_second": 0.212,
|
| 120 |
+
"learning_rate": 0.001,
|
| 121 |
+
"step": 1911
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 7.326007326007326,
|
| 125 |
+
"grad_norm": 0.22214488685131073,
|
| 126 |
+
"learning_rate": 0.001,
|
| 127 |
+
"loss": 0.1763,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.0,
|
| 132 |
+
"eval_accuracy": 0.24185724185724186,
|
| 133 |
+
"eval_f1_macro": 0.587992292024695,
|
| 134 |
+
"eval_f1_micro": 0.7655172413793103,
|
| 135 |
+
"eval_loss": 0.15311872959136963,
|
| 136 |
+
"eval_roc_auc": 0.837740052624858,
|
| 137 |
+
"eval_runtime": 427.9308,
|
| 138 |
+
"eval_samples_per_second": 6.744,
|
| 139 |
+
"eval_steps_per_second": 0.213,
|
| 140 |
+
"learning_rate": 0.001,
|
| 141 |
+
"step": 2184
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 9.0,
|
| 145 |
+
"eval_accuracy": 0.24012474012474014,
|
| 146 |
+
"eval_f1_macro": 0.606908576330327,
|
| 147 |
+
"eval_f1_micro": 0.7699542669773061,
|
| 148 |
+
"eval_loss": 0.14992575347423553,
|
| 149 |
+
"eval_roc_auc": 0.8431046707780733,
|
| 150 |
+
"eval_runtime": 424.0382,
|
| 151 |
+
"eval_samples_per_second": 6.806,
|
| 152 |
+
"eval_steps_per_second": 0.215,
|
| 153 |
+
"learning_rate": 0.001,
|
| 154 |
+
"step": 2457
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 9.157509157509157,
|
| 158 |
+
"grad_norm": 0.1733015924692154,
|
| 159 |
+
"learning_rate": 0.001,
|
| 160 |
+
"loss": 0.1735,
|
| 161 |
+
"step": 2500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 10.0,
|
| 165 |
+
"eval_accuracy": 0.24393624393624394,
|
| 166 |
+
"eval_f1_macro": 0.5829080312220596,
|
| 167 |
+
"eval_f1_micro": 0.7606115107913669,
|
| 168 |
+
"eval_loss": 0.1509619951248169,
|
| 169 |
+
"eval_roc_auc": 0.8277441062627229,
|
| 170 |
+
"eval_runtime": 424.8811,
|
| 171 |
+
"eval_samples_per_second": 6.792,
|
| 172 |
+
"eval_steps_per_second": 0.214,
|
| 173 |
+
"learning_rate": 0.001,
|
| 174 |
+
"step": 2730
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 10.989010989010989,
|
| 178 |
+
"grad_norm": 0.16356830298900604,
|
| 179 |
+
"learning_rate": 0.001,
|
| 180 |
+
"loss": 0.1723,
|
| 181 |
+
"step": 3000
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 11.0,
|
| 185 |
+
"eval_accuracy": 0.2505197505197505,
|
| 186 |
+
"eval_f1_macro": 0.5976223089766404,
|
| 187 |
+
"eval_f1_micro": 0.7689559002963221,
|
| 188 |
+
"eval_loss": 0.1520717293024063,
|
| 189 |
+
"eval_roc_auc": 0.8399853012032679,
|
| 190 |
+
"eval_runtime": 434.5331,
|
| 191 |
+
"eval_samples_per_second": 6.642,
|
| 192 |
+
"eval_steps_per_second": 0.209,
|
| 193 |
+
"learning_rate": 0.001,
|
| 194 |
+
"step": 3003
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 12.0,
|
| 198 |
+
"eval_accuracy": 0.2442827442827443,
|
| 199 |
+
"eval_f1_macro": 0.607405900640871,
|
| 200 |
+
"eval_f1_micro": 0.7759986516096409,
|
| 201 |
+
"eval_loss": 0.15027731657028198,
|
| 202 |
+
"eval_roc_auc": 0.8526551998703694,
|
| 203 |
+
"eval_runtime": 434.0545,
|
| 204 |
+
"eval_samples_per_second": 6.649,
|
| 205 |
+
"eval_steps_per_second": 0.21,
|
| 206 |
+
"learning_rate": 0.001,
|
| 207 |
+
"step": 3276
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 12.820512820512821,
|
| 211 |
+
"grad_norm": 0.1642971783876419,
|
| 212 |
+
"learning_rate": 0.001,
|
| 213 |
+
"loss": 0.1719,
|
| 214 |
+
"step": 3500
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 13.0,
|
| 218 |
+
"eval_accuracy": 0.24393624393624394,
|
| 219 |
+
"eval_f1_macro": 0.6003271512523337,
|
| 220 |
+
"eval_f1_micro": 0.7623558852444365,
|
| 221 |
+
"eval_loss": 0.1504218876361847,
|
| 222 |
+
"eval_roc_auc": 0.8301696089299148,
|
| 223 |
+
"eval_runtime": 426.4716,
|
| 224 |
+
"eval_samples_per_second": 6.767,
|
| 225 |
+
"eval_steps_per_second": 0.213,
|
| 226 |
+
"learning_rate": 0.001,
|
| 227 |
+
"step": 3549
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 14.0,
|
| 231 |
+
"eval_accuracy": 0.24462924462924462,
|
| 232 |
+
"eval_f1_macro": 0.602811285040826,
|
| 233 |
+
"eval_f1_micro": 0.7644358114073813,
|
| 234 |
+
"eval_loss": 0.1496724784374237,
|
| 235 |
+
"eval_roc_auc": 0.8342951177137805,
|
| 236 |
+
"eval_runtime": 428.909,
|
| 237 |
+
"eval_samples_per_second": 6.729,
|
| 238 |
+
"eval_steps_per_second": 0.212,
|
| 239 |
+
"learning_rate": 0.001,
|
| 240 |
+
"step": 3822
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 14.652014652014651,
|
| 244 |
+
"grad_norm": 0.1759812980890274,
|
| 245 |
+
"learning_rate": 0.001,
|
| 246 |
+
"loss": 0.1702,
|
| 247 |
+
"step": 4000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 15.0,
|
| 251 |
+
"eval_accuracy": 0.2512127512127512,
|
| 252 |
+
"eval_f1_macro": 0.6066013767027806,
|
| 253 |
+
"eval_f1_micro": 0.7751615281210703,
|
| 254 |
+
"eval_loss": 0.14749661087989807,
|
| 255 |
+
"eval_roc_auc": 0.8445581856657356,
|
| 256 |
+
"eval_runtime": 424.6732,
|
| 257 |
+
"eval_samples_per_second": 6.796,
|
| 258 |
+
"eval_steps_per_second": 0.214,
|
| 259 |
+
"learning_rate": 0.001,
|
| 260 |
+
"step": 4095
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 16.0,
|
| 264 |
+
"eval_accuracy": 0.24636174636174638,
|
| 265 |
+
"eval_f1_macro": 0.5838354990739413,
|
| 266 |
+
"eval_f1_micro": 0.7645565108923241,
|
| 267 |
+
"eval_loss": 0.14998775720596313,
|
| 268 |
+
"eval_roc_auc": 0.8320747114163963,
|
| 269 |
+
"eval_runtime": 423.7704,
|
| 270 |
+
"eval_samples_per_second": 6.81,
|
| 271 |
+
"eval_steps_per_second": 0.215,
|
| 272 |
+
"learning_rate": 0.001,
|
| 273 |
+
"step": 4368
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 16.483516483516482,
|
| 277 |
+
"grad_norm": 0.14804692566394806,
|
| 278 |
+
"learning_rate": 0.001,
|
| 279 |
+
"loss": 0.1696,
|
| 280 |
+
"step": 4500
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 17.0,
|
| 284 |
+
"eval_accuracy": 0.24566874566874566,
|
| 285 |
+
"eval_f1_macro": 0.6073459016890155,
|
| 286 |
+
"eval_f1_micro": 0.7719883641341547,
|
| 287 |
+
"eval_loss": 0.15297245979309082,
|
| 288 |
+
"eval_roc_auc": 0.8464322218871764,
|
| 289 |
+
"eval_runtime": 424.9885,
|
| 290 |
+
"eval_samples_per_second": 6.791,
|
| 291 |
+
"eval_steps_per_second": 0.214,
|
| 292 |
+
"learning_rate": 0.001,
|
| 293 |
+
"step": 4641
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 18.0,
|
| 297 |
+
"eval_accuracy": 0.24393624393624394,
|
| 298 |
+
"eval_f1_macro": 0.614324753279198,
|
| 299 |
+
"eval_f1_micro": 0.7751951282271207,
|
| 300 |
+
"eval_loss": 0.14907290041446686,
|
| 301 |
+
"eval_roc_auc": 0.8475019020709771,
|
| 302 |
+
"eval_runtime": 420.1647,
|
| 303 |
+
"eval_samples_per_second": 6.869,
|
| 304 |
+
"eval_steps_per_second": 0.217,
|
| 305 |
+
"learning_rate": 0.001,
|
| 306 |
+
"step": 4914
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 18.315018315018314,
|
| 310 |
+
"grad_norm": 0.19223743677139282,
|
| 311 |
+
"learning_rate": 0.001,
|
| 312 |
+
"loss": 0.1717,
|
| 313 |
+
"step": 5000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 19.0,
|
| 317 |
+
"eval_accuracy": 0.23458073458073458,
|
| 318 |
+
"eval_f1_macro": 0.6075499214740471,
|
| 319 |
+
"eval_f1_micro": 0.7739734788726388,
|
| 320 |
+
"eval_loss": 0.14951026439666748,
|
| 321 |
+
"eval_roc_auc": 0.848377592477135,
|
| 322 |
+
"eval_runtime": 427.9682,
|
| 323 |
+
"eval_samples_per_second": 6.743,
|
| 324 |
+
"eval_steps_per_second": 0.213,
|
| 325 |
+
"learning_rate": 0.001,
|
| 326 |
+
"step": 5187
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 20.0,
|
| 330 |
+
"eval_accuracy": 0.24532224532224534,
|
| 331 |
+
"eval_f1_macro": 0.595638442008225,
|
| 332 |
+
"eval_f1_micro": 0.7636993911381718,
|
| 333 |
+
"eval_loss": 0.14873762428760529,
|
| 334 |
+
"eval_roc_auc": 0.8322311292560515,
|
| 335 |
+
"eval_runtime": 421.5059,
|
| 336 |
+
"eval_samples_per_second": 6.847,
|
| 337 |
+
"eval_steps_per_second": 0.216,
|
| 338 |
+
"learning_rate": 0.001,
|
| 339 |
+
"step": 5460
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 20.146520146520146,
|
| 343 |
+
"grad_norm": 0.15787707269191742,
|
| 344 |
+
"learning_rate": 0.001,
|
| 345 |
+
"loss": 0.1705,
|
| 346 |
+
"step": 5500
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 21.0,
|
| 350 |
+
"eval_accuracy": 0.24740124740124741,
|
| 351 |
+
"eval_f1_macro": 0.6164990545073296,
|
| 352 |
+
"eval_f1_micro": 0.780452718426063,
|
| 353 |
+
"eval_loss": 0.14705629646778107,
|
| 354 |
+
"eval_roc_auc": 0.8539786012990958,
|
| 355 |
+
"eval_runtime": 429.6596,
|
| 356 |
+
"eval_samples_per_second": 6.717,
|
| 357 |
+
"eval_steps_per_second": 0.212,
|
| 358 |
+
"learning_rate": 0.001,
|
| 359 |
+
"step": 5733
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 21.978021978021978,
|
| 363 |
+
"grad_norm": 0.15392103791236877,
|
| 364 |
+
"learning_rate": 0.001,
|
| 365 |
+
"loss": 0.1706,
|
| 366 |
+
"step": 6000
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 22.0,
|
| 370 |
+
"eval_accuracy": 0.24532224532224534,
|
| 371 |
+
"eval_f1_macro": 0.6073576225776433,
|
| 372 |
+
"eval_f1_micro": 0.7753641707130079,
|
| 373 |
+
"eval_loss": 0.1508719027042389,
|
| 374 |
+
"eval_roc_auc": 0.8494150259851333,
|
| 375 |
+
"eval_runtime": 429.7216,
|
| 376 |
+
"eval_samples_per_second": 6.716,
|
| 377 |
+
"eval_steps_per_second": 0.212,
|
| 378 |
+
"learning_rate": 0.001,
|
| 379 |
+
"step": 6006
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 23.0,
|
| 383 |
+
"eval_accuracy": 0.2428967428967429,
|
| 384 |
+
"eval_f1_macro": 0.6127152502703448,
|
| 385 |
+
"eval_f1_micro": 0.771920553133395,
|
| 386 |
+
"eval_loss": 0.15015815198421478,
|
| 387 |
+
"eval_roc_auc": 0.8388299205154317,
|
| 388 |
+
"eval_runtime": 426.6602,
|
| 389 |
+
"eval_samples_per_second": 6.764,
|
| 390 |
+
"eval_steps_per_second": 0.213,
|
| 391 |
+
"learning_rate": 0.001,
|
| 392 |
+
"step": 6279
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 23.80952380952381,
|
| 396 |
+
"grad_norm": 0.1737624853849411,
|
| 397 |
+
"learning_rate": 0.001,
|
| 398 |
+
"loss": 0.1699,
|
| 399 |
+
"step": 6500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 24.0,
|
| 403 |
+
"eval_accuracy": 0.24012474012474014,
|
| 404 |
+
"eval_f1_macro": 0.5849380548549015,
|
| 405 |
+
"eval_f1_micro": 0.7698941591532732,
|
| 406 |
+
"eval_loss": 0.14965225756168365,
|
| 407 |
+
"eval_roc_auc": 0.8406060899537385,
|
| 408 |
+
"eval_runtime": 430.4521,
|
| 409 |
+
"eval_samples_per_second": 6.705,
|
| 410 |
+
"eval_steps_per_second": 0.211,
|
| 411 |
+
"learning_rate": 0.001,
|
| 412 |
+
"step": 6552
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 25.0,
|
| 416 |
+
"eval_accuracy": 0.24255024255024255,
|
| 417 |
+
"eval_f1_macro": 0.6035289549510865,
|
| 418 |
+
"eval_f1_micro": 0.7761348897535668,
|
| 419 |
+
"eval_loss": 0.14702074229717255,
|
| 420 |
+
"eval_roc_auc": 0.8458632504863829,
|
| 421 |
+
"eval_runtime": 428.0693,
|
| 422 |
+
"eval_samples_per_second": 6.742,
|
| 423 |
+
"eval_steps_per_second": 0.213,
|
| 424 |
+
"learning_rate": 0.001,
|
| 425 |
+
"step": 6825
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 25.641025641025642,
|
| 429 |
+
"grad_norm": 0.1737377792596817,
|
| 430 |
+
"learning_rate": 0.001,
|
| 431 |
+
"loss": 0.1694,
|
| 432 |
+
"step": 7000
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 26.0,
|
| 436 |
+
"eval_accuracy": 0.24220374220374222,
|
| 437 |
+
"eval_f1_macro": 0.6064603919289959,
|
| 438 |
+
"eval_f1_micro": 0.7751430907604253,
|
| 439 |
+
"eval_loss": 0.14808295667171478,
|
| 440 |
+
"eval_roc_auc": 0.8465518457868458,
|
| 441 |
+
"eval_runtime": 438.4341,
|
| 442 |
+
"eval_samples_per_second": 6.583,
|
| 443 |
+
"eval_steps_per_second": 0.208,
|
| 444 |
+
"learning_rate": 0.001,
|
| 445 |
+
"step": 7098
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 27.0,
|
| 449 |
+
"eval_accuracy": 0.24740124740124741,
|
| 450 |
+
"eval_f1_macro": 0.6135774018658996,
|
| 451 |
+
"eval_f1_micro": 0.7689308343302761,
|
| 452 |
+
"eval_loss": 0.14581289887428284,
|
| 453 |
+
"eval_roc_auc": 0.8357120666953542,
|
| 454 |
+
"eval_runtime": 426.6923,
|
| 455 |
+
"eval_samples_per_second": 6.764,
|
| 456 |
+
"eval_steps_per_second": 0.213,
|
| 457 |
+
"learning_rate": 0.001,
|
| 458 |
+
"step": 7371
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 27.47252747252747,
|
| 462 |
+
"grad_norm": 0.16500511765480042,
|
| 463 |
+
"learning_rate": 0.001,
|
| 464 |
+
"loss": 0.17,
|
| 465 |
+
"step": 7500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 28.0,
|
| 469 |
+
"eval_accuracy": 0.24462924462924462,
|
| 470 |
+
"eval_f1_macro": 0.6077297645661711,
|
| 471 |
+
"eval_f1_micro": 0.7751325049960902,
|
| 472 |
+
"eval_loss": 0.1453842669725418,
|
| 473 |
+
"eval_roc_auc": 0.8440532649625113,
|
| 474 |
+
"eval_runtime": 431.4145,
|
| 475 |
+
"eval_samples_per_second": 6.69,
|
| 476 |
+
"eval_steps_per_second": 0.211,
|
| 477 |
+
"learning_rate": 0.001,
|
| 478 |
+
"step": 7644
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 29.0,
|
| 482 |
+
"eval_accuracy": 0.24566874566874566,
|
| 483 |
+
"eval_f1_macro": 0.6107922701154117,
|
| 484 |
+
"eval_f1_micro": 0.7735191637630662,
|
| 485 |
+
"eval_loss": 0.14941243827342987,
|
| 486 |
+
"eval_roc_auc": 0.849050708300112,
|
| 487 |
+
"eval_runtime": 434.9588,
|
| 488 |
+
"eval_samples_per_second": 6.635,
|
| 489 |
+
"eval_steps_per_second": 0.209,
|
| 490 |
+
"learning_rate": 0.001,
|
| 491 |
+
"step": 7917
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 29.304029304029303,
|
| 495 |
+
"grad_norm": 0.1599486619234085,
|
| 496 |
+
"learning_rate": 0.001,
|
| 497 |
+
"loss": 0.1685,
|
| 498 |
+
"step": 8000
|
| 499 |
+
},
|
| 500 |
+
{
|
| 501 |
+
"epoch": 30.0,
|
| 502 |
+
"eval_accuracy": 0.24982674982674982,
|
| 503 |
+
"eval_f1_macro": 0.5982833860845571,
|
| 504 |
+
"eval_f1_micro": 0.7705324709843182,
|
| 505 |
+
"eval_loss": 0.14549985527992249,
|
| 506 |
+
"eval_roc_auc": 0.8366026732011344,
|
| 507 |
+
"eval_runtime": 434.3329,
|
| 508 |
+
"eval_samples_per_second": 6.645,
|
| 509 |
+
"eval_steps_per_second": 0.21,
|
| 510 |
+
"learning_rate": 0.001,
|
| 511 |
+
"step": 8190
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 31.0,
|
| 515 |
+
"eval_accuracy": 0.2532917532917533,
|
| 516 |
+
"eval_f1_macro": 0.6068619458731248,
|
| 517 |
+
"eval_f1_micro": 0.7784728768532008,
|
| 518 |
+
"eval_loss": 0.14541107416152954,
|
| 519 |
+
"eval_roc_auc": 0.8494949988142239,
|
| 520 |
+
"eval_runtime": 435.6219,
|
| 521 |
+
"eval_samples_per_second": 6.625,
|
| 522 |
+
"eval_steps_per_second": 0.209,
|
| 523 |
+
"learning_rate": 0.001,
|
| 524 |
+
"step": 8463
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 31.135531135531135,
|
| 528 |
+
"grad_norm": 0.1950293928384781,
|
| 529 |
+
"learning_rate": 0.001,
|
| 530 |
+
"loss": 0.1687,
|
| 531 |
+
"step": 8500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 32.0,
|
| 535 |
+
"eval_accuracy": 0.24532224532224534,
|
| 536 |
+
"eval_f1_macro": 0.6145316287096297,
|
| 537 |
+
"eval_f1_micro": 0.7746102833519939,
|
| 538 |
+
"eval_loss": 0.14657220244407654,
|
| 539 |
+
"eval_roc_auc": 0.8460955499587395,
|
| 540 |
+
"eval_runtime": 434.8949,
|
| 541 |
+
"eval_samples_per_second": 6.636,
|
| 542 |
+
"eval_steps_per_second": 0.209,
|
| 543 |
+
"learning_rate": 0.001,
|
| 544 |
+
"step": 8736
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 32.967032967032964,
|
| 548 |
+
"grad_norm": 0.18405263125896454,
|
| 549 |
+
"learning_rate": 0.001,
|
| 550 |
+
"loss": 0.1679,
|
| 551 |
+
"step": 9000
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 33.0,
|
| 555 |
+
"eval_accuracy": 0.253984753984754,
|
| 556 |
+
"eval_f1_macro": 0.6124691593400795,
|
| 557 |
+
"eval_f1_micro": 0.777031154551008,
|
| 558 |
+
"eval_loss": 0.14459234476089478,
|
| 559 |
+
"eval_roc_auc": 0.843919167617255,
|
| 560 |
+
"eval_runtime": 440.1591,
|
| 561 |
+
"eval_samples_per_second": 6.557,
|
| 562 |
+
"eval_steps_per_second": 0.207,
|
| 563 |
+
"learning_rate": 0.001,
|
| 564 |
+
"step": 9009
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 34.0,
|
| 568 |
+
"eval_accuracy": 0.24462924462924462,
|
| 569 |
+
"eval_f1_macro": 0.6168054796129936,
|
| 570 |
+
"eval_f1_micro": 0.7781283769180896,
|
| 571 |
+
"eval_loss": 0.1468168944120407,
|
| 572 |
+
"eval_roc_auc": 0.8469846407097918,
|
| 573 |
+
"eval_runtime": 438.6105,
|
| 574 |
+
"eval_samples_per_second": 6.58,
|
| 575 |
+
"eval_steps_per_second": 0.207,
|
| 576 |
+
"learning_rate": 0.001,
|
| 577 |
+
"step": 9282
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 34.798534798534796,
|
| 581 |
+
"grad_norm": 0.17146140336990356,
|
| 582 |
+
"learning_rate": 0.001,
|
| 583 |
+
"loss": 0.168,
|
| 584 |
+
"step": 9500
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 35.0,
|
| 588 |
+
"eval_accuracy": 0.2494802494802495,
|
| 589 |
+
"eval_f1_macro": 0.6193343400891848,
|
| 590 |
+
"eval_f1_micro": 0.7766880749869814,
|
| 591 |
+
"eval_loss": 0.14858707785606384,
|
| 592 |
+
"eval_roc_auc": 0.8451765062846143,
|
| 593 |
+
"eval_runtime": 434.5802,
|
| 594 |
+
"eval_samples_per_second": 6.641,
|
| 595 |
+
"eval_steps_per_second": 0.209,
|
| 596 |
+
"learning_rate": 0.001,
|
| 597 |
+
"step": 9555
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 36.0,
|
| 601 |
+
"eval_accuracy": 0.24878724878724878,
|
| 602 |
+
"eval_f1_macro": 0.6092667253949349,
|
| 603 |
+
"eval_f1_micro": 0.7718835224773468,
|
| 604 |
+
"eval_loss": 0.14637114107608795,
|
| 605 |
+
"eval_roc_auc": 0.8391158347811251,
|
| 606 |
+
"eval_runtime": 439.3197,
|
| 607 |
+
"eval_samples_per_second": 6.569,
|
| 608 |
+
"eval_steps_per_second": 0.207,
|
| 609 |
+
"learning_rate": 0.001,
|
| 610 |
+
"step": 9828
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 36.63003663003663,
|
| 614 |
+
"grad_norm": 0.16876503825187683,
|
| 615 |
+
"learning_rate": 0.001,
|
| 616 |
+
"loss": 0.169,
|
| 617 |
+
"step": 10000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 37.0,
|
| 621 |
+
"eval_accuracy": 0.24982674982674982,
|
| 622 |
+
"eval_f1_macro": 0.6127183895875491,
|
| 623 |
+
"eval_f1_micro": 0.7733602776435442,
|
| 624 |
+
"eval_loss": 0.1448281705379486,
|
| 625 |
+
"eval_roc_auc": 0.8402195590843876,
|
| 626 |
+
"eval_runtime": 437.3035,
|
| 627 |
+
"eval_samples_per_second": 6.6,
|
| 628 |
+
"eval_steps_per_second": 0.208,
|
| 629 |
+
"learning_rate": 0.001,
|
| 630 |
+
"step": 10101
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 38.0,
|
| 634 |
+
"eval_accuracy": 0.25225225225225223,
|
| 635 |
+
"eval_f1_macro": 0.6109962510638844,
|
| 636 |
+
"eval_f1_micro": 0.7814896880859042,
|
| 637 |
+
"eval_loss": 0.1450735628604889,
|
| 638 |
+
"eval_roc_auc": 0.8526187412743501,
|
| 639 |
+
"eval_runtime": 437.7229,
|
| 640 |
+
"eval_samples_per_second": 6.593,
|
| 641 |
+
"eval_steps_per_second": 0.208,
|
| 642 |
+
"learning_rate": 0.001,
|
| 643 |
+
"step": 10374
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 38.46153846153846,
|
| 647 |
+
"grad_norm": 0.19475676119327545,
|
| 648 |
+
"learning_rate": 0.001,
|
| 649 |
+
"loss": 0.167,
|
| 650 |
+
"step": 10500
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 39.0,
|
| 654 |
+
"eval_accuracy": 0.24982674982674982,
|
| 655 |
+
"eval_f1_macro": 0.6272196317832909,
|
| 656 |
+
"eval_f1_micro": 0.7824146207942057,
|
| 657 |
+
"eval_loss": 0.14469724893569946,
|
| 658 |
+
"eval_roc_auc": 0.8563424677452759,
|
| 659 |
+
"eval_runtime": 435.4486,
|
| 660 |
+
"eval_samples_per_second": 6.628,
|
| 661 |
+
"eval_steps_per_second": 0.209,
|
| 662 |
+
"learning_rate": 0.001,
|
| 663 |
+
"step": 10647
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 40.0,
|
| 667 |
+
"eval_accuracy": 0.25363825363825365,
|
| 668 |
+
"eval_f1_macro": 0.6265963634718456,
|
| 669 |
+
"eval_f1_micro": 0.7836651178652115,
|
| 670 |
+
"eval_loss": 0.14824891090393066,
|
| 671 |
+
"eval_roc_auc": 0.853692740688437,
|
| 672 |
+
"eval_runtime": 435.8824,
|
| 673 |
+
"eval_samples_per_second": 6.621,
|
| 674 |
+
"eval_steps_per_second": 0.209,
|
| 675 |
+
"learning_rate": 0.0001,
|
| 676 |
+
"step": 10920
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 40.29304029304029,
|
| 680 |
+
"grad_norm": 0.15533967316150665,
|
| 681 |
+
"learning_rate": 0.0001,
|
| 682 |
+
"loss": 0.1652,
|
| 683 |
+
"step": 11000
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 41.0,
|
| 687 |
+
"eval_accuracy": 0.2616077616077616,
|
| 688 |
+
"eval_f1_macro": 0.6323784470247855,
|
| 689 |
+
"eval_f1_micro": 0.7833456473553827,
|
| 690 |
+
"eval_loss": 0.14141727983951569,
|
| 691 |
+
"eval_roc_auc": 0.8483120796798727,
|
| 692 |
+
"eval_runtime": 435.7344,
|
| 693 |
+
"eval_samples_per_second": 6.623,
|
| 694 |
+
"eval_steps_per_second": 0.209,
|
| 695 |
+
"learning_rate": 0.0001,
|
| 696 |
+
"step": 11193
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 42.0,
|
| 700 |
+
"eval_accuracy": 0.26195426195426197,
|
| 701 |
+
"eval_f1_macro": 0.6371841233046203,
|
| 702 |
+
"eval_f1_micro": 0.7884351407000686,
|
| 703 |
+
"eval_loss": 0.13979895412921906,
|
| 704 |
+
"eval_roc_auc": 0.8545567611245666,
|
| 705 |
+
"eval_runtime": 438.4508,
|
| 706 |
+
"eval_samples_per_second": 6.582,
|
| 707 |
+
"eval_steps_per_second": 0.208,
|
| 708 |
+
"learning_rate": 0.0001,
|
| 709 |
+
"step": 11466
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 42.124542124542124,
|
| 713 |
+
"grad_norm": 0.1733330935239792,
|
| 714 |
+
"learning_rate": 0.0001,
|
| 715 |
+
"loss": 0.1608,
|
| 716 |
+
"step": 11500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 43.0,
|
| 720 |
+
"eval_accuracy": 0.26403326403326405,
|
| 721 |
+
"eval_f1_macro": 0.6366820358518588,
|
| 722 |
+
"eval_f1_micro": 0.7871061893724783,
|
| 723 |
+
"eval_loss": 0.14107641577720642,
|
| 724 |
+
"eval_roc_auc": 0.853678548931782,
|
| 725 |
+
"eval_runtime": 434.1211,
|
| 726 |
+
"eval_samples_per_second": 6.648,
|
| 727 |
+
"eval_steps_per_second": 0.21,
|
| 728 |
+
"learning_rate": 0.0001,
|
| 729 |
+
"step": 11739
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 43.956043956043956,
|
| 733 |
+
"grad_norm": 0.19694675505161285,
|
| 734 |
+
"learning_rate": 0.0001,
|
| 735 |
+
"loss": 0.1596,
|
| 736 |
+
"step": 12000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 44.0,
|
| 740 |
+
"eval_accuracy": 0.26126126126126126,
|
| 741 |
+
"eval_f1_macro": 0.6256922069455233,
|
| 742 |
+
"eval_f1_micro": 0.787878787878788,
|
| 743 |
+
"eval_loss": 0.13898694515228271,
|
| 744 |
+
"eval_roc_auc": 0.8537086091649239,
|
| 745 |
+
"eval_runtime": 434.0073,
|
| 746 |
+
"eval_samples_per_second": 6.65,
|
| 747 |
+
"eval_steps_per_second": 0.21,
|
| 748 |
+
"learning_rate": 0.0001,
|
| 749 |
+
"step": 12012
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 45.0,
|
| 753 |
+
"eval_accuracy": 0.2664587664587665,
|
| 754 |
+
"eval_f1_macro": 0.6421056073559387,
|
| 755 |
+
"eval_f1_micro": 0.7894011202068074,
|
| 756 |
+
"eval_loss": 0.13859130442142487,
|
| 757 |
+
"eval_roc_auc": 0.8538817942028954,
|
| 758 |
+
"eval_runtime": 432.4865,
|
| 759 |
+
"eval_samples_per_second": 6.673,
|
| 760 |
+
"eval_steps_per_second": 0.21,
|
| 761 |
+
"learning_rate": 0.0001,
|
| 762 |
+
"step": 12285
|
| 763 |
+
},
|
| 764 |
+
{
|
| 765 |
+
"epoch": 45.78754578754579,
|
| 766 |
+
"grad_norm": 0.18810147047042847,
|
| 767 |
+
"learning_rate": 0.0001,
|
| 768 |
+
"loss": 0.1582,
|
| 769 |
+
"step": 12500
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 46.0,
|
| 773 |
+
"eval_accuracy": 0.2664587664587665,
|
| 774 |
+
"eval_f1_macro": 0.6283048537279357,
|
| 775 |
+
"eval_f1_micro": 0.7873893327575039,
|
| 776 |
+
"eval_loss": 0.139601469039917,
|
| 777 |
+
"eval_roc_auc": 0.8521625527563127,
|
| 778 |
+
"eval_runtime": 421.9429,
|
| 779 |
+
"eval_samples_per_second": 6.84,
|
| 780 |
+
"eval_steps_per_second": 0.216,
|
| 781 |
+
"learning_rate": 0.0001,
|
| 782 |
+
"step": 12558
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 47.0,
|
| 786 |
+
"eval_accuracy": 0.2636867636867637,
|
| 787 |
+
"eval_f1_macro": 0.6286555138094179,
|
| 788 |
+
"eval_f1_micro": 0.7863567238757333,
|
| 789 |
+
"eval_loss": 0.13869330286979675,
|
| 790 |
+
"eval_roc_auc": 0.8499808451526433,
|
| 791 |
+
"eval_runtime": 424.0306,
|
| 792 |
+
"eval_samples_per_second": 6.806,
|
| 793 |
+
"eval_steps_per_second": 0.215,
|
| 794 |
+
"learning_rate": 0.0001,
|
| 795 |
+
"step": 12831
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 47.61904761904762,
|
| 799 |
+
"grad_norm": 0.15351006388664246,
|
| 800 |
+
"learning_rate": 0.0001,
|
| 801 |
+
"loss": 0.1584,
|
| 802 |
+
"step": 13000
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 48.0,
|
| 806 |
+
"eval_accuracy": 0.26784476784476785,
|
| 807 |
+
"eval_f1_macro": 0.6334934953582803,
|
| 808 |
+
"eval_f1_micro": 0.7913177234660741,
|
| 809 |
+
"eval_loss": 0.13777127861976624,
|
| 810 |
+
"eval_roc_auc": 0.8571892112602602,
|
| 811 |
+
"eval_runtime": 419.9652,
|
| 812 |
+
"eval_samples_per_second": 6.872,
|
| 813 |
+
"eval_steps_per_second": 0.217,
|
| 814 |
+
"learning_rate": 0.0001,
|
| 815 |
+
"step": 13104
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 49.0,
|
| 819 |
+
"eval_accuracy": 0.26403326403326405,
|
| 820 |
+
"eval_f1_macro": 0.6381777921693204,
|
| 821 |
+
"eval_f1_micro": 0.7933989479042932,
|
| 822 |
+
"eval_loss": 0.1377096027135849,
|
| 823 |
+
"eval_roc_auc": 0.8602965218660363,
|
| 824 |
+
"eval_runtime": 431.2306,
|
| 825 |
+
"eval_samples_per_second": 6.692,
|
| 826 |
+
"eval_steps_per_second": 0.211,
|
| 827 |
+
"learning_rate": 0.0001,
|
| 828 |
+
"step": 13377
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 49.45054945054945,
|
| 832 |
+
"grad_norm": 0.1798904836177826,
|
| 833 |
+
"learning_rate": 0.0001,
|
| 834 |
+
"loss": 0.157,
|
| 835 |
+
"step": 13500
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 50.0,
|
| 839 |
+
"eval_accuracy": 0.2674982674982675,
|
| 840 |
+
"eval_f1_macro": 0.6362718007605523,
|
| 841 |
+
"eval_f1_micro": 0.7918342891380639,
|
| 842 |
+
"eval_loss": 0.13755330443382263,
|
| 843 |
+
"eval_roc_auc": 0.8570210161405075,
|
| 844 |
+
"eval_runtime": 429.5809,
|
| 845 |
+
"eval_samples_per_second": 6.718,
|
| 846 |
+
"eval_steps_per_second": 0.212,
|
| 847 |
+
"learning_rate": 0.0001,
|
| 848 |
+
"step": 13650
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 51.0,
|
| 852 |
+
"eval_accuracy": 0.2661122661122661,
|
| 853 |
+
"eval_f1_macro": 0.6426825970872383,
|
| 854 |
+
"eval_f1_micro": 0.7928808087673094,
|
| 855 |
+
"eval_loss": 0.13754987716674805,
|
| 856 |
+
"eval_roc_auc": 0.8596608706709776,
|
| 857 |
+
"eval_runtime": 429.3766,
|
| 858 |
+
"eval_samples_per_second": 6.721,
|
| 859 |
+
"eval_steps_per_second": 0.212,
|
| 860 |
+
"learning_rate": 0.0001,
|
| 861 |
+
"step": 13923
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 51.282051282051285,
|
| 865 |
+
"grad_norm": 0.20376506447792053,
|
| 866 |
+
"learning_rate": 0.0001,
|
| 867 |
+
"loss": 0.1567,
|
| 868 |
+
"step": 14000
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 52.0,
|
| 872 |
+
"eval_accuracy": 0.26576576576576577,
|
| 873 |
+
"eval_f1_macro": 0.6367912909960436,
|
| 874 |
+
"eval_f1_micro": 0.7871186146434616,
|
| 875 |
+
"eval_loss": 0.13771678507328033,
|
| 876 |
+
"eval_roc_auc": 0.8506886757830149,
|
| 877 |
+
"eval_runtime": 424.3804,
|
| 878 |
+
"eval_samples_per_second": 6.801,
|
| 879 |
+
"eval_steps_per_second": 0.214,
|
| 880 |
+
"learning_rate": 0.0001,
|
| 881 |
+
"step": 14196
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 53.0,
|
| 885 |
+
"eval_accuracy": 0.2692307692307692,
|
| 886 |
+
"eval_f1_macro": 0.640555047060403,
|
| 887 |
+
"eval_f1_micro": 0.7928592630284527,
|
| 888 |
+
"eval_loss": 0.13740690052509308,
|
| 889 |
+
"eval_roc_auc": 0.8601326459765699,
|
| 890 |
+
"eval_runtime": 434.4832,
|
| 891 |
+
"eval_samples_per_second": 6.642,
|
| 892 |
+
"eval_steps_per_second": 0.209,
|
| 893 |
+
"learning_rate": 0.0001,
|
| 894 |
+
"step": 14469
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 53.11355311355312,
|
| 898 |
+
"grad_norm": 0.16348811984062195,
|
| 899 |
+
"learning_rate": 0.0001,
|
| 900 |
+
"loss": 0.1571,
|
| 901 |
+
"step": 14500
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 54.0,
|
| 905 |
+
"eval_accuracy": 0.27165627165627165,
|
| 906 |
+
"eval_f1_macro": 0.6412320555565514,
|
| 907 |
+
"eval_f1_micro": 0.7920979171140219,
|
| 908 |
+
"eval_loss": 0.1368684023618698,
|
| 909 |
+
"eval_roc_auc": 0.8562094300869534,
|
| 910 |
+
"eval_runtime": 425.2932,
|
| 911 |
+
"eval_samples_per_second": 6.786,
|
| 912 |
+
"eval_steps_per_second": 0.214,
|
| 913 |
+
"learning_rate": 0.0001,
|
| 914 |
+
"step": 14742
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 54.94505494505494,
|
| 918 |
+
"grad_norm": 0.20431332290172577,
|
| 919 |
+
"learning_rate": 0.0001,
|
| 920 |
+
"loss": 0.1548,
|
| 921 |
+
"step": 15000
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 55.0,
|
| 925 |
+
"eval_accuracy": 0.2702702702702703,
|
| 926 |
+
"eval_f1_macro": 0.6377616721633446,
|
| 927 |
+
"eval_f1_micro": 0.7914089347079037,
|
| 928 |
+
"eval_loss": 0.13703426718711853,
|
| 929 |
+
"eval_roc_auc": 0.8557803910164303,
|
| 930 |
+
"eval_runtime": 424.9893,
|
| 931 |
+
"eval_samples_per_second": 6.791,
|
| 932 |
+
"eval_steps_per_second": 0.214,
|
| 933 |
+
"learning_rate": 0.0001,
|
| 934 |
+
"step": 15015
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 56.0,
|
| 938 |
+
"eval_accuracy": 0.2643797643797644,
|
| 939 |
+
"eval_f1_macro": 0.6425003998141597,
|
| 940 |
+
"eval_f1_micro": 0.7931107623128156,
|
| 941 |
+
"eval_loss": 0.1364637017250061,
|
| 942 |
+
"eval_roc_auc": 0.8601515459625123,
|
| 943 |
+
"eval_runtime": 423.7139,
|
| 944 |
+
"eval_samples_per_second": 6.811,
|
| 945 |
+
"eval_steps_per_second": 0.215,
|
| 946 |
+
"learning_rate": 0.0001,
|
| 947 |
+
"step": 15288
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 56.776556776556774,
|
| 951 |
+
"grad_norm": 0.19714656472206116,
|
| 952 |
+
"learning_rate": 0.0001,
|
| 953 |
+
"loss": 0.155,
|
| 954 |
+
"step": 15500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 57.0,
|
| 958 |
+
"eval_accuracy": 0.2674982674982675,
|
| 959 |
+
"eval_f1_macro": 0.6381793578718891,
|
| 960 |
+
"eval_f1_micro": 0.7926408585665006,
|
| 961 |
+
"eval_loss": 0.13675515353679657,
|
| 962 |
+
"eval_roc_auc": 0.8588114846455387,
|
| 963 |
+
"eval_runtime": 426.4919,
|
| 964 |
+
"eval_samples_per_second": 6.767,
|
| 965 |
+
"eval_steps_per_second": 0.213,
|
| 966 |
+
"learning_rate": 0.0001,
|
| 967 |
+
"step": 15561
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 58.0,
|
| 971 |
+
"eval_accuracy": 0.2674982674982675,
|
| 972 |
+
"eval_f1_macro": 0.637380953089336,
|
| 973 |
+
"eval_f1_micro": 0.791562634524322,
|
| 974 |
+
"eval_loss": 0.1364695280790329,
|
| 975 |
+
"eval_roc_auc": 0.855274853280308,
|
| 976 |
+
"eval_runtime": 425.8426,
|
| 977 |
+
"eval_samples_per_second": 6.777,
|
| 978 |
+
"eval_steps_per_second": 0.214,
|
| 979 |
+
"learning_rate": 0.0001,
|
| 980 |
+
"step": 15834
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 58.608058608058606,
|
| 984 |
+
"grad_norm": 0.19042669236660004,
|
| 985 |
+
"learning_rate": 0.0001,
|
| 986 |
+
"loss": 0.155,
|
| 987 |
+
"step": 16000
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 59.0,
|
| 991 |
+
"eval_accuracy": 0.2674982674982675,
|
| 992 |
+
"eval_f1_macro": 0.6428884521567982,
|
| 993 |
+
"eval_f1_micro": 0.7922245108135942,
|
| 994 |
+
"eval_loss": 0.13641765713691711,
|
| 995 |
+
"eval_roc_auc": 0.8565012329926954,
|
| 996 |
+
"eval_runtime": 423.8693,
|
| 997 |
+
"eval_samples_per_second": 6.809,
|
| 998 |
+
"eval_steps_per_second": 0.215,
|
| 999 |
+
"learning_rate": 0.0001,
|
| 1000 |
+
"step": 16107
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 60.0,
|
| 1004 |
+
"eval_accuracy": 0.26507276507276506,
|
| 1005 |
+
"eval_f1_macro": 0.6357999016219877,
|
| 1006 |
+
"eval_f1_micro": 0.7882888744307093,
|
| 1007 |
+
"eval_loss": 0.13687649369239807,
|
| 1008 |
+
"eval_roc_auc": 0.8514745744887481,
|
| 1009 |
+
"eval_runtime": 423.4928,
|
| 1010 |
+
"eval_samples_per_second": 6.815,
|
| 1011 |
+
"eval_steps_per_second": 0.215,
|
| 1012 |
+
"learning_rate": 0.0001,
|
| 1013 |
+
"step": 16380
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 60.43956043956044,
|
| 1017 |
+
"grad_norm": 0.18568764626979828,
|
| 1018 |
+
"learning_rate": 0.0001,
|
| 1019 |
+
"loss": 0.1546,
|
| 1020 |
+
"step": 16500
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 61.0,
|
| 1024 |
+
"eval_accuracy": 0.2713097713097713,
|
| 1025 |
+
"eval_f1_macro": 0.6503848519713329,
|
| 1026 |
+
"eval_f1_micro": 0.7945638702508654,
|
| 1027 |
+
"eval_loss": 0.13638463616371155,
|
| 1028 |
+
"eval_roc_auc": 0.8588833823919201,
|
| 1029 |
+
"eval_runtime": 425.9119,
|
| 1030 |
+
"eval_samples_per_second": 6.776,
|
| 1031 |
+
"eval_steps_per_second": 0.214,
|
| 1032 |
+
"learning_rate": 0.0001,
|
| 1033 |
+
"step": 16653
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 62.0,
|
| 1037 |
+
"eval_accuracy": 0.2751212751212751,
|
| 1038 |
+
"eval_f1_macro": 0.6441767594174573,
|
| 1039 |
+
"eval_f1_micro": 0.7931640039405492,
|
| 1040 |
+
"eval_loss": 0.13563227653503418,
|
| 1041 |
+
"eval_roc_auc": 0.8575138778747027,
|
| 1042 |
+
"eval_runtime": 422.0661,
|
| 1043 |
+
"eval_samples_per_second": 6.838,
|
| 1044 |
+
"eval_steps_per_second": 0.216,
|
| 1045 |
+
"learning_rate": 0.0001,
|
| 1046 |
+
"step": 16926
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 62.27106227106227,
|
| 1050 |
+
"grad_norm": 0.19402863085269928,
|
| 1051 |
+
"learning_rate": 0.0001,
|
| 1052 |
+
"loss": 0.1536,
|
| 1053 |
+
"step": 17000
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 63.0,
|
| 1057 |
+
"eval_accuracy": 0.27373527373527373,
|
| 1058 |
+
"eval_f1_macro": 0.6515952055035917,
|
| 1059 |
+
"eval_f1_micro": 0.7966116124638174,
|
| 1060 |
+
"eval_loss": 0.1355270892381668,
|
| 1061 |
+
"eval_roc_auc": 0.8610939161629354,
|
| 1062 |
+
"eval_runtime": 426.9279,
|
| 1063 |
+
"eval_samples_per_second": 6.76,
|
| 1064 |
+
"eval_steps_per_second": 0.213,
|
| 1065 |
+
"learning_rate": 0.0001,
|
| 1066 |
+
"step": 17199
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 64.0,
|
| 1070 |
+
"eval_accuracy": 0.26784476784476785,
|
| 1071 |
+
"eval_f1_macro": 0.6450040026439422,
|
| 1072 |
+
"eval_f1_micro": 0.7934075342465754,
|
| 1073 |
+
"eval_loss": 0.13592010736465454,
|
| 1074 |
+
"eval_roc_auc": 0.8577985580745997,
|
| 1075 |
+
"eval_runtime": 426.0816,
|
| 1076 |
+
"eval_samples_per_second": 6.773,
|
| 1077 |
+
"eval_steps_per_second": 0.214,
|
| 1078 |
+
"learning_rate": 0.0001,
|
| 1079 |
+
"step": 17472
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 64.1025641025641,
|
| 1083 |
+
"grad_norm": 0.22000150382518768,
|
| 1084 |
+
"learning_rate": 0.0001,
|
| 1085 |
+
"loss": 0.1544,
|
| 1086 |
+
"step": 17500
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 65.0,
|
| 1090 |
+
"eval_accuracy": 0.27061677061677064,
|
| 1091 |
+
"eval_f1_macro": 0.64551501310817,
|
| 1092 |
+
"eval_f1_micro": 0.7936467053015668,
|
| 1093 |
+
"eval_loss": 0.13569533824920654,
|
| 1094 |
+
"eval_roc_auc": 0.857159821715051,
|
| 1095 |
+
"eval_runtime": 424.6551,
|
| 1096 |
+
"eval_samples_per_second": 6.796,
|
| 1097 |
+
"eval_steps_per_second": 0.214,
|
| 1098 |
+
"learning_rate": 0.0001,
|
| 1099 |
+
"step": 17745
|
| 1100 |
+
},
|
| 1101 |
+
{
|
| 1102 |
+
"epoch": 65.93406593406593,
|
| 1103 |
+
"grad_norm": 0.19799016416072845,
|
| 1104 |
+
"learning_rate": 0.0001,
|
| 1105 |
+
"loss": 0.1529,
|
| 1106 |
+
"step": 18000
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 66.0,
|
| 1110 |
+
"eval_accuracy": 0.2713097713097713,
|
| 1111 |
+
"eval_f1_macro": 0.6477176853690674,
|
| 1112 |
+
"eval_f1_micro": 0.794643237940888,
|
| 1113 |
+
"eval_loss": 0.13565082848072052,
|
| 1114 |
+
"eval_roc_auc": 0.8594942449609874,
|
| 1115 |
+
"eval_runtime": 425.0795,
|
| 1116 |
+
"eval_samples_per_second": 6.789,
|
| 1117 |
+
"eval_steps_per_second": 0.214,
|
| 1118 |
+
"learning_rate": 0.0001,
|
| 1119 |
+
"step": 18018
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 67.0,
|
| 1123 |
+
"eval_accuracy": 0.27546777546777546,
|
| 1124 |
+
"eval_f1_macro": 0.6544361257862924,
|
| 1125 |
+
"eval_f1_micro": 0.7965922095536813,
|
| 1126 |
+
"eval_loss": 0.13533934950828552,
|
| 1127 |
+
"eval_roc_auc": 0.8622831129363361,
|
| 1128 |
+
"eval_runtime": 424.6762,
|
| 1129 |
+
"eval_samples_per_second": 6.796,
|
| 1130 |
+
"eval_steps_per_second": 0.214,
|
| 1131 |
+
"learning_rate": 0.0001,
|
| 1132 |
+
"step": 18291
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 67.76556776556777,
|
| 1136 |
+
"grad_norm": 0.2619948983192444,
|
| 1137 |
+
"learning_rate": 0.0001,
|
| 1138 |
+
"loss": 0.1528,
|
| 1139 |
+
"step": 18500
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 68.0,
|
| 1143 |
+
"eval_accuracy": 0.2733887733887734,
|
| 1144 |
+
"eval_f1_macro": 0.6519486064773884,
|
| 1145 |
+
"eval_f1_micro": 0.7955772910907932,
|
| 1146 |
+
"eval_loss": 0.1353396475315094,
|
| 1147 |
+
"eval_roc_auc": 0.8608058154545816,
|
| 1148 |
+
"eval_runtime": 421.8067,
|
| 1149 |
+
"eval_samples_per_second": 6.842,
|
| 1150 |
+
"eval_steps_per_second": 0.216,
|
| 1151 |
+
"learning_rate": 0.0001,
|
| 1152 |
+
"step": 18564
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 69.0,
|
| 1156 |
+
"eval_accuracy": 0.26992376992376993,
|
| 1157 |
+
"eval_f1_macro": 0.6515714856354324,
|
| 1158 |
+
"eval_f1_micro": 0.7966188524590164,
|
| 1159 |
+
"eval_loss": 0.13474246859550476,
|
| 1160 |
+
"eval_roc_auc": 0.8602900698481241,
|
| 1161 |
+
"eval_runtime": 423.2901,
|
| 1162 |
+
"eval_samples_per_second": 6.818,
|
| 1163 |
+
"eval_steps_per_second": 0.215,
|
| 1164 |
+
"learning_rate": 0.0001,
|
| 1165 |
+
"step": 18837
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 69.59706959706959,
|
| 1169 |
+
"grad_norm": 0.18048201501369476,
|
| 1170 |
+
"learning_rate": 0.0001,
|
| 1171 |
+
"loss": 0.1528,
|
| 1172 |
+
"step": 19000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 70.0,
|
| 1176 |
+
"eval_accuracy": 0.272002772002772,
|
| 1177 |
+
"eval_f1_macro": 0.6441608871918139,
|
| 1178 |
+
"eval_f1_micro": 0.7944687795241776,
|
| 1179 |
+
"eval_loss": 0.13504748046398163,
|
| 1180 |
+
"eval_roc_auc": 0.8574953132327267,
|
| 1181 |
+
"eval_runtime": 423.3844,
|
| 1182 |
+
"eval_samples_per_second": 6.817,
|
| 1183 |
+
"eval_steps_per_second": 0.215,
|
| 1184 |
+
"learning_rate": 0.0001,
|
| 1185 |
+
"step": 19110
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 71.0,
|
| 1189 |
+
"eval_accuracy": 0.27234927234927236,
|
| 1190 |
+
"eval_f1_macro": 0.6441889860402124,
|
| 1191 |
+
"eval_f1_micro": 0.7933057280883367,
|
| 1192 |
+
"eval_loss": 0.13502468168735504,
|
| 1193 |
+
"eval_roc_auc": 0.8556664277229126,
|
| 1194 |
+
"eval_runtime": 422.6912,
|
| 1195 |
+
"eval_samples_per_second": 6.828,
|
| 1196 |
+
"eval_steps_per_second": 0.215,
|
| 1197 |
+
"learning_rate": 0.0001,
|
| 1198 |
+
"step": 19383
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 71.42857142857143,
|
| 1202 |
+
"grad_norm": 0.24162879586219788,
|
| 1203 |
+
"learning_rate": 0.0001,
|
| 1204 |
+
"loss": 0.1522,
|
| 1205 |
+
"step": 19500
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 72.0,
|
| 1209 |
+
"eval_accuracy": 0.2758142758142758,
|
| 1210 |
+
"eval_f1_macro": 0.6484748365424647,
|
| 1211 |
+
"eval_f1_micro": 0.7969950486597234,
|
| 1212 |
+
"eval_loss": 0.1344645917415619,
|
| 1213 |
+
"eval_roc_auc": 0.8605409876174911,
|
| 1214 |
+
"eval_runtime": 426.5755,
|
| 1215 |
+
"eval_samples_per_second": 6.766,
|
| 1216 |
+
"eval_steps_per_second": 0.213,
|
| 1217 |
+
"learning_rate": 0.0001,
|
| 1218 |
+
"step": 19656
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 73.0,
|
| 1222 |
+
"eval_accuracy": 0.27616077616077617,
|
| 1223 |
+
"eval_f1_macro": 0.6518769914193778,
|
| 1224 |
+
"eval_f1_micro": 0.7977006599957419,
|
| 1225 |
+
"eval_loss": 0.1341526359319687,
|
| 1226 |
+
"eval_roc_auc": 0.8616010233088203,
|
| 1227 |
+
"eval_runtime": 420.7226,
|
| 1228 |
+
"eval_samples_per_second": 6.86,
|
| 1229 |
+
"eval_steps_per_second": 0.216,
|
| 1230 |
+
"learning_rate": 0.0001,
|
| 1231 |
+
"step": 19929
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 73.26007326007326,
|
| 1235 |
+
"grad_norm": 0.22451983392238617,
|
| 1236 |
+
"learning_rate": 0.0001,
|
| 1237 |
+
"loss": 0.1523,
|
| 1238 |
+
"step": 20000
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 74.0,
|
| 1242 |
+
"eval_accuracy": 0.2751212751212751,
|
| 1243 |
+
"eval_f1_macro": 0.641334935505441,
|
| 1244 |
+
"eval_f1_micro": 0.7914797229603171,
|
| 1245 |
+
"eval_loss": 0.13499116897583008,
|
| 1246 |
+
"eval_roc_auc": 0.8520198169504839,
|
| 1247 |
+
"eval_runtime": 428.7922,
|
| 1248 |
+
"eval_samples_per_second": 6.731,
|
| 1249 |
+
"eval_steps_per_second": 0.212,
|
| 1250 |
+
"learning_rate": 0.0001,
|
| 1251 |
+
"step": 20202
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 75.0,
|
| 1255 |
+
"eval_accuracy": 0.2751212751212751,
|
| 1256 |
+
"eval_f1_macro": 0.6485229770180625,
|
| 1257 |
+
"eval_f1_micro": 0.7946678133734681,
|
| 1258 |
+
"eval_loss": 0.13461369276046753,
|
| 1259 |
+
"eval_roc_auc": 0.8572354216588205,
|
| 1260 |
+
"eval_runtime": 427.8784,
|
| 1261 |
+
"eval_samples_per_second": 6.745,
|
| 1262 |
+
"eval_steps_per_second": 0.213,
|
| 1263 |
+
"learning_rate": 0.0001,
|
| 1264 |
+
"step": 20475
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 75.0915750915751,
|
| 1268 |
+
"grad_norm": 0.22029711306095123,
|
| 1269 |
+
"learning_rate": 0.0001,
|
| 1270 |
+
"loss": 0.1521,
|
| 1271 |
+
"step": 20500
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 76.0,
|
| 1275 |
+
"eval_accuracy": 0.2758142758142758,
|
| 1276 |
+
"eval_f1_macro": 0.6478195810395848,
|
| 1277 |
+
"eval_f1_micro": 0.7964594201659113,
|
| 1278 |
+
"eval_loss": 0.13438266515731812,
|
| 1279 |
+
"eval_roc_auc": 0.8597526207801657,
|
| 1280 |
+
"eval_runtime": 424.3142,
|
| 1281 |
+
"eval_samples_per_second": 6.802,
|
| 1282 |
+
"eval_steps_per_second": 0.214,
|
| 1283 |
+
"learning_rate": 0.0001,
|
| 1284 |
+
"step": 20748
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 76.92307692307692,
|
| 1288 |
+
"grad_norm": 0.2415299415588379,
|
| 1289 |
+
"learning_rate": 0.0001,
|
| 1290 |
+
"loss": 0.1515,
|
| 1291 |
+
"step": 21000
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 77.0,
|
| 1295 |
+
"eval_accuracy": 0.27754677754677753,
|
| 1296 |
+
"eval_f1_macro": 0.6536737916153181,
|
| 1297 |
+
"eval_f1_micro": 0.7977742853502102,
|
| 1298 |
+
"eval_loss": 0.13460540771484375,
|
| 1299 |
+
"eval_roc_auc": 0.8623314561225224,
|
| 1300 |
+
"eval_runtime": 422.8083,
|
| 1301 |
+
"eval_samples_per_second": 6.826,
|
| 1302 |
+
"eval_steps_per_second": 0.215,
|
| 1303 |
+
"learning_rate": 0.0001,
|
| 1304 |
+
"step": 21021
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 78.0,
|
| 1308 |
+
"eval_accuracy": 0.27754677754677753,
|
| 1309 |
+
"eval_f1_macro": 0.6543115985953537,
|
| 1310 |
+
"eval_f1_micro": 0.7978169818504888,
|
| 1311 |
+
"eval_loss": 0.13411369919776917,
|
| 1312 |
+
"eval_roc_auc": 0.8634738791194995,
|
| 1313 |
+
"eval_runtime": 428.5067,
|
| 1314 |
+
"eval_samples_per_second": 6.735,
|
| 1315 |
+
"eval_steps_per_second": 0.212,
|
| 1316 |
+
"learning_rate": 0.0001,
|
| 1317 |
+
"step": 21294
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 78.75457875457876,
|
| 1321 |
+
"grad_norm": 0.2636328637599945,
|
| 1322 |
+
"learning_rate": 0.0001,
|
| 1323 |
+
"loss": 0.1514,
|
| 1324 |
+
"step": 21500
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 79.0,
|
| 1328 |
+
"eval_accuracy": 0.2740817740817741,
|
| 1329 |
+
"eval_f1_macro": 0.6523004018612216,
|
| 1330 |
+
"eval_f1_micro": 0.7953020134228188,
|
| 1331 |
+
"eval_loss": 0.13399606943130493,
|
| 1332 |
+
"eval_roc_auc": 0.8574454542918126,
|
| 1333 |
+
"eval_runtime": 436.7976,
|
| 1334 |
+
"eval_samples_per_second": 6.607,
|
| 1335 |
+
"eval_steps_per_second": 0.208,
|
| 1336 |
+
"learning_rate": 0.0001,
|
| 1337 |
+
"step": 21567
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 80.0,
|
| 1341 |
+
"eval_accuracy": 0.27823977823977825,
|
| 1342 |
+
"eval_f1_macro": 0.6545582038870168,
|
| 1343 |
+
"eval_f1_micro": 0.7993085420355848,
|
| 1344 |
+
"eval_loss": 0.1344238668680191,
|
| 1345 |
+
"eval_roc_auc": 0.8652547567870936,
|
| 1346 |
+
"eval_runtime": 431.9941,
|
| 1347 |
+
"eval_samples_per_second": 6.681,
|
| 1348 |
+
"eval_steps_per_second": 0.211,
|
| 1349 |
+
"learning_rate": 0.0001,
|
| 1350 |
+
"step": 21840
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 80.58608058608058,
|
| 1354 |
+
"grad_norm": 0.23601791262626648,
|
| 1355 |
+
"learning_rate": 0.0001,
|
| 1356 |
+
"loss": 0.1516,
|
| 1357 |
+
"step": 22000
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 81.0,
|
| 1361 |
+
"eval_accuracy": 0.2758142758142758,
|
| 1362 |
+
"eval_f1_macro": 0.6559691700651434,
|
| 1363 |
+
"eval_f1_micro": 0.7966715529878418,
|
| 1364 |
+
"eval_loss": 0.13405664265155792,
|
| 1365 |
+
"eval_roc_auc": 0.8575861109650502,
|
| 1366 |
+
"eval_runtime": 436.6356,
|
| 1367 |
+
"eval_samples_per_second": 6.61,
|
| 1368 |
+
"eval_steps_per_second": 0.208,
|
| 1369 |
+
"learning_rate": 0.0001,
|
| 1370 |
+
"step": 22113
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 82.0,
|
| 1374 |
+
"eval_accuracy": 0.2765072765072765,
|
| 1375 |
+
"eval_f1_macro": 0.6453669674995801,
|
| 1376 |
+
"eval_f1_micro": 0.7947541551246537,
|
| 1377 |
+
"eval_loss": 0.13407430052757263,
|
| 1378 |
+
"eval_roc_auc": 0.8554945304057716,
|
| 1379 |
+
"eval_runtime": 436.5794,
|
| 1380 |
+
"eval_samples_per_second": 6.61,
|
| 1381 |
+
"eval_steps_per_second": 0.208,
|
| 1382 |
+
"learning_rate": 0.0001,
|
| 1383 |
+
"step": 22386
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 82.41758241758242,
|
| 1387 |
+
"grad_norm": 0.19588124752044678,
|
| 1388 |
+
"learning_rate": 0.0001,
|
| 1389 |
+
"loss": 0.149,
|
| 1390 |
+
"step": 22500
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 83.0,
|
| 1394 |
+
"eval_accuracy": 0.2702702702702703,
|
| 1395 |
+
"eval_f1_macro": 0.645966570658811,
|
| 1396 |
+
"eval_f1_micro": 0.7924365020985678,
|
| 1397 |
+
"eval_loss": 0.1350804716348648,
|
| 1398 |
+
"eval_roc_auc": 0.8543412288505268,
|
| 1399 |
+
"eval_runtime": 433.6987,
|
| 1400 |
+
"eval_samples_per_second": 6.654,
|
| 1401 |
+
"eval_steps_per_second": 0.21,
|
| 1402 |
+
"learning_rate": 0.0001,
|
| 1403 |
+
"step": 22659
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 84.0,
|
| 1407 |
+
"eval_accuracy": 0.27546777546777546,
|
| 1408 |
+
"eval_f1_macro": 0.6512285101875886,
|
| 1409 |
+
"eval_f1_micro": 0.7957293542577825,
|
| 1410 |
+
"eval_loss": 0.13387472927570343,
|
| 1411 |
+
"eval_roc_auc": 0.8585996545688873,
|
| 1412 |
+
"eval_runtime": 432.4386,
|
| 1413 |
+
"eval_samples_per_second": 6.674,
|
| 1414 |
+
"eval_steps_per_second": 0.21,
|
| 1415 |
+
"learning_rate": 0.0001,
|
| 1416 |
+
"step": 22932
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 84.24908424908425,
|
| 1420 |
+
"grad_norm": 0.2560372054576874,
|
| 1421 |
+
"learning_rate": 0.0001,
|
| 1422 |
+
"loss": 0.1515,
|
| 1423 |
+
"step": 23000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 85.0,
|
| 1427 |
+
"eval_accuracy": 0.27927927927927926,
|
| 1428 |
+
"eval_f1_macro": 0.6531817491521362,
|
| 1429 |
+
"eval_f1_micro": 0.7990622335890879,
|
| 1430 |
+
"eval_loss": 0.13341927528381348,
|
| 1431 |
+
"eval_roc_auc": 0.8620406055936447,
|
| 1432 |
+
"eval_runtime": 432.3488,
|
| 1433 |
+
"eval_samples_per_second": 6.675,
|
| 1434 |
+
"eval_steps_per_second": 0.21,
|
| 1435 |
+
"learning_rate": 0.0001,
|
| 1436 |
+
"step": 23205
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 86.0,
|
| 1440 |
+
"eval_accuracy": 0.2747747747747748,
|
| 1441 |
+
"eval_f1_macro": 0.6595866427349153,
|
| 1442 |
+
"eval_f1_micro": 0.7988261313371896,
|
| 1443 |
+
"eval_loss": 0.13337253034114838,
|
| 1444 |
+
"eval_roc_auc": 0.8625331319838734,
|
| 1445 |
+
"eval_runtime": 435.2436,
|
| 1446 |
+
"eval_samples_per_second": 6.631,
|
| 1447 |
+
"eval_steps_per_second": 0.209,
|
| 1448 |
+
"learning_rate": 0.0001,
|
| 1449 |
+
"step": 23478
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 86.08058608058609,
|
| 1453 |
+
"grad_norm": 0.28640052676200867,
|
| 1454 |
+
"learning_rate": 0.0001,
|
| 1455 |
+
"loss": 0.1495,
|
| 1456 |
+
"step": 23500
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 87.0,
|
| 1460 |
+
"eval_accuracy": 0.27442827442827444,
|
| 1461 |
+
"eval_f1_macro": 0.6467323251879672,
|
| 1462 |
+
"eval_f1_micro": 0.7956179390619651,
|
| 1463 |
+
"eval_loss": 0.1339845359325409,
|
| 1464 |
+
"eval_roc_auc": 0.8590850582532711,
|
| 1465 |
+
"eval_runtime": 438.7375,
|
| 1466 |
+
"eval_samples_per_second": 6.578,
|
| 1467 |
+
"eval_steps_per_second": 0.207,
|
| 1468 |
+
"learning_rate": 0.0001,
|
| 1469 |
+
"step": 23751
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 87.91208791208791,
|
| 1473 |
+
"grad_norm": 0.23546907305717468,
|
| 1474 |
+
"learning_rate": 0.0001,
|
| 1475 |
+
"loss": 0.1496,
|
| 1476 |
+
"step": 24000
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 88.0,
|
| 1480 |
+
"eval_accuracy": 0.2747747747747748,
|
| 1481 |
+
"eval_f1_macro": 0.648318545746826,
|
| 1482 |
+
"eval_f1_micro": 0.7981612326551459,
|
| 1483 |
+
"eval_loss": 0.13357459008693695,
|
| 1484 |
+
"eval_roc_auc": 0.8619578829440303,
|
| 1485 |
+
"eval_runtime": 432.3449,
|
| 1486 |
+
"eval_samples_per_second": 6.675,
|
| 1487 |
+
"eval_steps_per_second": 0.21,
|
| 1488 |
+
"learning_rate": 0.0001,
|
| 1489 |
+
"step": 24024
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 89.0,
|
| 1493 |
+
"eval_accuracy": 0.2806652806652807,
|
| 1494 |
+
"eval_f1_macro": 0.6585340844298272,
|
| 1495 |
+
"eval_f1_micro": 0.8014968675104065,
|
| 1496 |
+
"eval_loss": 0.13366733491420746,
|
| 1497 |
+
"eval_roc_auc": 0.8672320387088881,
|
| 1498 |
+
"eval_runtime": 431.6296,
|
| 1499 |
+
"eval_samples_per_second": 6.686,
|
| 1500 |
+
"eval_steps_per_second": 0.211,
|
| 1501 |
+
"learning_rate": 0.0001,
|
| 1502 |
+
"step": 24297
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 89.74358974358974,
|
| 1506 |
+
"grad_norm": 0.24246211349964142,
|
| 1507 |
+
"learning_rate": 0.0001,
|
| 1508 |
+
"loss": 0.1493,
|
| 1509 |
+
"step": 24500
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 90.0,
|
| 1513 |
+
"eval_accuracy": 0.2772002772002772,
|
| 1514 |
+
"eval_f1_macro": 0.66211749340029,
|
| 1515 |
+
"eval_f1_micro": 0.8010798042854732,
|
| 1516 |
+
"eval_loss": 0.1332736760377884,
|
| 1517 |
+
"eval_roc_auc": 0.8661044781564988,
|
| 1518 |
+
"eval_runtime": 425.5723,
|
| 1519 |
+
"eval_samples_per_second": 6.781,
|
| 1520 |
+
"eval_steps_per_second": 0.214,
|
| 1521 |
+
"learning_rate": 0.0001,
|
| 1522 |
+
"step": 24570
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 91.0,
|
| 1526 |
+
"eval_accuracy": 0.27823977823977825,
|
| 1527 |
+
"eval_f1_macro": 0.6528573832362276,
|
| 1528 |
+
"eval_f1_micro": 0.7956933454403943,
|
| 1529 |
+
"eval_loss": 0.13367226719856262,
|
| 1530 |
+
"eval_roc_auc": 0.8562680347985093,
|
| 1531 |
+
"eval_runtime": 443.8961,
|
| 1532 |
+
"eval_samples_per_second": 6.502,
|
| 1533 |
+
"eval_steps_per_second": 0.205,
|
| 1534 |
+
"learning_rate": 0.0001,
|
| 1535 |
+
"step": 24843
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 91.57509157509158,
|
| 1539 |
+
"grad_norm": 0.22026851773262024,
|
| 1540 |
+
"learning_rate": 0.0001,
|
| 1541 |
+
"loss": 0.1496,
|
| 1542 |
+
"step": 25000
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 92.0,
|
| 1546 |
+
"eval_accuracy": 0.27546777546777546,
|
| 1547 |
+
"eval_f1_macro": 0.6513649424471982,
|
| 1548 |
+
"eval_f1_micro": 0.796086375587259,
|
| 1549 |
+
"eval_loss": 0.13348612189292908,
|
| 1550 |
+
"eval_roc_auc": 0.8573559442803198,
|
| 1551 |
+
"eval_runtime": 443.9031,
|
| 1552 |
+
"eval_samples_per_second": 6.501,
|
| 1553 |
+
"eval_steps_per_second": 0.205,
|
| 1554 |
+
"learning_rate": 0.0001,
|
| 1555 |
+
"step": 25116
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 93.0,
|
| 1559 |
+
"eval_accuracy": 0.2758142758142758,
|
| 1560 |
+
"eval_f1_macro": 0.6559763883082907,
|
| 1561 |
+
"eval_f1_micro": 0.8001861094662043,
|
| 1562 |
+
"eval_loss": 0.1330718696117401,
|
| 1563 |
+
"eval_roc_auc": 0.8648260530605368,
|
| 1564 |
+
"eval_runtime": 436.5725,
|
| 1565 |
+
"eval_samples_per_second": 6.611,
|
| 1566 |
+
"eval_steps_per_second": 0.208,
|
| 1567 |
+
"learning_rate": 0.0001,
|
| 1568 |
+
"step": 25389
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 93.4065934065934,
|
| 1572 |
+
"grad_norm": 0.28630152344703674,
|
| 1573 |
+
"learning_rate": 0.0001,
|
| 1574 |
+
"loss": 0.1493,
|
| 1575 |
+
"step": 25500
|
| 1576 |
+
},
|
| 1577 |
+
{
|
| 1578 |
+
"epoch": 94.0,
|
| 1579 |
+
"eval_accuracy": 0.2758142758142758,
|
| 1580 |
+
"eval_f1_macro": 0.6553585917255438,
|
| 1581 |
+
"eval_f1_micro": 0.7995090362720617,
|
| 1582 |
+
"eval_loss": 0.13329002261161804,
|
| 1583 |
+
"eval_roc_auc": 0.864277443745379,
|
| 1584 |
+
"eval_runtime": 442.8808,
|
| 1585 |
+
"eval_samples_per_second": 6.516,
|
| 1586 |
+
"eval_steps_per_second": 0.205,
|
| 1587 |
+
"learning_rate": 0.0001,
|
| 1588 |
+
"step": 25662
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 95.0,
|
| 1592 |
+
"eval_accuracy": 0.2758142758142758,
|
| 1593 |
+
"eval_f1_macro": 0.6579543710907207,
|
| 1594 |
+
"eval_f1_micro": 0.7979651162790697,
|
| 1595 |
+
"eval_loss": 0.13314621150493622,
|
| 1596 |
+
"eval_roc_auc": 0.8606367216129991,
|
| 1597 |
+
"eval_runtime": 436.3942,
|
| 1598 |
+
"eval_samples_per_second": 6.613,
|
| 1599 |
+
"eval_steps_per_second": 0.209,
|
| 1600 |
+
"learning_rate": 0.0001,
|
| 1601 |
+
"step": 25935
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 95.23809523809524,
|
| 1605 |
+
"grad_norm": 0.25194719433784485,
|
| 1606 |
+
"learning_rate": 0.0001,
|
| 1607 |
+
"loss": 0.1482,
|
| 1608 |
+
"step": 26000
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 96.0,
|
| 1612 |
+
"eval_accuracy": 0.2751212751212751,
|
| 1613 |
+
"eval_f1_macro": 0.6556445954379041,
|
| 1614 |
+
"eval_f1_micro": 0.7992523999660183,
|
| 1615 |
+
"eval_loss": 0.13279949128627777,
|
| 1616 |
+
"eval_roc_auc": 0.8631226264354063,
|
| 1617 |
+
"eval_runtime": 426.8086,
|
| 1618 |
+
"eval_samples_per_second": 6.762,
|
| 1619 |
+
"eval_steps_per_second": 0.213,
|
| 1620 |
+
"learning_rate": 0.0001,
|
| 1621 |
+
"step": 26208
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 97.0,
|
| 1625 |
+
"eval_accuracy": 0.27823977823977825,
|
| 1626 |
+
"eval_f1_macro": 0.6492741904723621,
|
| 1627 |
+
"eval_f1_micro": 0.7977296181630549,
|
| 1628 |
+
"eval_loss": 0.1332886964082718,
|
| 1629 |
+
"eval_roc_auc": 0.8588905587527994,
|
| 1630 |
+
"eval_runtime": 441.9848,
|
| 1631 |
+
"eval_samples_per_second": 6.53,
|
| 1632 |
+
"eval_steps_per_second": 0.206,
|
| 1633 |
+
"learning_rate": 0.0001,
|
| 1634 |
+
"step": 26481
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 97.06959706959707,
|
| 1638 |
+
"grad_norm": 0.27280953526496887,
|
| 1639 |
+
"learning_rate": 0.0001,
|
| 1640 |
+
"loss": 0.1497,
|
| 1641 |
+
"step": 26500
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 98.0,
|
| 1645 |
+
"eval_accuracy": 0.27546777546777546,
|
| 1646 |
+
"eval_f1_macro": 0.6600105762308898,
|
| 1647 |
+
"eval_f1_micro": 0.799611141637432,
|
| 1648 |
+
"eval_loss": 0.13266970217227936,
|
| 1649 |
+
"eval_roc_auc": 0.864715456620441,
|
| 1650 |
+
"eval_runtime": 439.781,
|
| 1651 |
+
"eval_samples_per_second": 6.562,
|
| 1652 |
+
"eval_steps_per_second": 0.207,
|
| 1653 |
+
"learning_rate": 0.0001,
|
| 1654 |
+
"step": 26754
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 98.9010989010989,
|
| 1658 |
+
"grad_norm": 0.30599892139434814,
|
| 1659 |
+
"learning_rate": 0.0001,
|
| 1660 |
+
"loss": 0.1489,
|
| 1661 |
+
"step": 27000
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 99.0,
|
| 1665 |
+
"eval_accuracy": 0.27165627165627165,
|
| 1666 |
+
"eval_f1_macro": 0.6589970862385839,
|
| 1667 |
+
"eval_f1_micro": 0.7978809757764771,
|
| 1668 |
+
"eval_loss": 0.13253149390220642,
|
| 1669 |
+
"eval_roc_auc": 0.8607699202364255,
|
| 1670 |
+
"eval_runtime": 438.5456,
|
| 1671 |
+
"eval_samples_per_second": 6.581,
|
| 1672 |
+
"eval_steps_per_second": 0.208,
|
| 1673 |
+
"learning_rate": 0.0001,
|
| 1674 |
+
"step": 27027
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 100.0,
|
| 1678 |
+
"eval_accuracy": 0.27616077616077617,
|
| 1679 |
+
"eval_f1_macro": 0.6570195655430786,
|
| 1680 |
+
"eval_f1_micro": 0.797143840330351,
|
| 1681 |
+
"eval_loss": 0.1329408884048462,
|
| 1682 |
+
"eval_roc_auc": 0.8584810367011169,
|
| 1683 |
+
"eval_runtime": 434.9771,
|
| 1684 |
+
"eval_samples_per_second": 6.635,
|
| 1685 |
+
"eval_steps_per_second": 0.209,
|
| 1686 |
+
"learning_rate": 0.0001,
|
| 1687 |
+
"step": 27300
|
| 1688 |
+
},
|
| 1689 |
+
{
|
| 1690 |
+
"epoch": 100.73260073260073,
|
| 1691 |
+
"grad_norm": 0.2732805013656616,
|
| 1692 |
+
"learning_rate": 0.0001,
|
| 1693 |
+
"loss": 0.1482,
|
| 1694 |
+
"step": 27500
|
| 1695 |
+
},
|
| 1696 |
+
{
|
| 1697 |
+
"epoch": 101.0,
|
| 1698 |
+
"eval_accuracy": 0.28205128205128205,
|
| 1699 |
+
"eval_f1_macro": 0.657951499975745,
|
| 1700 |
+
"eval_f1_micro": 0.7991615690636095,
|
| 1701 |
+
"eval_loss": 0.13274870812892914,
|
| 1702 |
+
"eval_roc_auc": 0.861103560655407,
|
| 1703 |
+
"eval_runtime": 435.4493,
|
| 1704 |
+
"eval_samples_per_second": 6.628,
|
| 1705 |
+
"eval_steps_per_second": 0.209,
|
| 1706 |
+
"learning_rate": 0.0001,
|
| 1707 |
+
"step": 27573
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 102.0,
|
| 1711 |
+
"eval_accuracy": 0.2817047817047817,
|
| 1712 |
+
"eval_f1_macro": 0.654306822863844,
|
| 1713 |
+
"eval_f1_micro": 0.7986821274228745,
|
| 1714 |
+
"eval_loss": 0.1326293796300888,
|
| 1715 |
+
"eval_roc_auc": 0.8607733407448822,
|
| 1716 |
+
"eval_runtime": 437.9645,
|
| 1717 |
+
"eval_samples_per_second": 6.59,
|
| 1718 |
+
"eval_steps_per_second": 0.208,
|
| 1719 |
+
"learning_rate": 0.0001,
|
| 1720 |
+
"step": 27846
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 102.56410256410257,
|
| 1724 |
+
"grad_norm": 0.23533137142658234,
|
| 1725 |
+
"learning_rate": 0.0001,
|
| 1726 |
+
"loss": 0.1474,
|
| 1727 |
+
"step": 28000
|
| 1728 |
+
},
|
| 1729 |
+
{
|
| 1730 |
+
"epoch": 103.0,
|
| 1731 |
+
"eval_accuracy": 0.2803187803187803,
|
| 1732 |
+
"eval_f1_macro": 0.6518495856500403,
|
| 1733 |
+
"eval_f1_micro": 0.7993688968487486,
|
| 1734 |
+
"eval_loss": 0.13247379660606384,
|
| 1735 |
+
"eval_roc_auc": 0.8620991566501659,
|
| 1736 |
+
"eval_runtime": 426.0566,
|
| 1737 |
+
"eval_samples_per_second": 6.774,
|
| 1738 |
+
"eval_steps_per_second": 0.214,
|
| 1739 |
+
"learning_rate": 0.0001,
|
| 1740 |
+
"step": 28119
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 104.0,
|
| 1744 |
+
"eval_accuracy": 0.27754677754677753,
|
| 1745 |
+
"eval_f1_macro": 0.6612536009112525,
|
| 1746 |
+
"eval_f1_micro": 0.8010850676047981,
|
| 1747 |
+
"eval_loss": 0.13315415382385254,
|
| 1748 |
+
"eval_roc_auc": 0.864729420343199,
|
| 1749 |
+
"eval_runtime": 425.2679,
|
| 1750 |
+
"eval_samples_per_second": 6.786,
|
| 1751 |
+
"eval_steps_per_second": 0.214,
|
| 1752 |
+
"learning_rate": 0.0001,
|
| 1753 |
+
"step": 28392
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 104.3956043956044,
|
| 1757 |
+
"grad_norm": 0.2809629738330841,
|
| 1758 |
+
"learning_rate": 0.0001,
|
| 1759 |
+
"loss": 0.1472,
|
| 1760 |
+
"step": 28500
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 105.0,
|
| 1764 |
+
"eval_accuracy": 0.2830907830907831,
|
| 1765 |
+
"eval_f1_macro": 0.6635718544409769,
|
| 1766 |
+
"eval_f1_micro": 0.8012698412698412,
|
| 1767 |
+
"eval_loss": 0.13218620419502258,
|
| 1768 |
+
"eval_roc_auc": 0.8652135899617869,
|
| 1769 |
+
"eval_runtime": 425.1586,
|
| 1770 |
+
"eval_samples_per_second": 6.788,
|
| 1771 |
+
"eval_steps_per_second": 0.214,
|
| 1772 |
+
"learning_rate": 0.0001,
|
| 1773 |
+
"step": 28665
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 106.0,
|
| 1777 |
+
"eval_accuracy": 0.2830907830907831,
|
| 1778 |
+
"eval_f1_macro": 0.6588128942023547,
|
| 1779 |
+
"eval_f1_micro": 0.800988243312319,
|
| 1780 |
+
"eval_loss": 0.13239973783493042,
|
| 1781 |
+
"eval_roc_auc": 0.8632750603887415,
|
| 1782 |
+
"eval_runtime": 427.5404,
|
| 1783 |
+
"eval_samples_per_second": 6.75,
|
| 1784 |
+
"eval_steps_per_second": 0.213,
|
| 1785 |
+
"learning_rate": 0.0001,
|
| 1786 |
+
"step": 28938
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 106.22710622710623,
|
| 1790 |
+
"grad_norm": 0.2568123936653137,
|
| 1791 |
+
"learning_rate": 0.0001,
|
| 1792 |
+
"loss": 0.148,
|
| 1793 |
+
"step": 29000
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 107.0,
|
| 1797 |
+
"eval_accuracy": 0.2785862785862786,
|
| 1798 |
+
"eval_f1_macro": 0.650564106362156,
|
| 1799 |
+
"eval_f1_micro": 0.7985513421389007,
|
| 1800 |
+
"eval_loss": 0.13358280062675476,
|
| 1801 |
+
"eval_roc_auc": 0.8618832353771251,
|
| 1802 |
+
"eval_runtime": 425.2874,
|
| 1803 |
+
"eval_samples_per_second": 6.786,
|
| 1804 |
+
"eval_steps_per_second": 0.214,
|
| 1805 |
+
"learning_rate": 0.0001,
|
| 1806 |
+
"step": 29211
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 108.0,
|
| 1810 |
+
"eval_accuracy": 0.2796257796257796,
|
| 1811 |
+
"eval_f1_macro": 0.6501303094783896,
|
| 1812 |
+
"eval_f1_micro": 0.7995554225623049,
|
| 1813 |
+
"eval_loss": 0.13270235061645508,
|
| 1814 |
+
"eval_roc_auc": 0.8615071940670409,
|
| 1815 |
+
"eval_runtime": 432.9179,
|
| 1816 |
+
"eval_samples_per_second": 6.666,
|
| 1817 |
+
"eval_steps_per_second": 0.21,
|
| 1818 |
+
"learning_rate": 0.0001,
|
| 1819 |
+
"step": 29484
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 108.05860805860806,
|
| 1823 |
+
"grad_norm": 0.29480934143066406,
|
| 1824 |
+
"learning_rate": 0.0001,
|
| 1825 |
+
"loss": 0.1477,
|
| 1826 |
+
"step": 29500
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 109.0,
|
| 1830 |
+
"eval_accuracy": 0.2806652806652807,
|
| 1831 |
+
"eval_f1_macro": 0.6579556871315007,
|
| 1832 |
+
"eval_f1_micro": 0.8000342553738118,
|
| 1833 |
+
"eval_loss": 0.1318453699350357,
|
| 1834 |
+
"eval_roc_auc": 0.8612993478767093,
|
| 1835 |
+
"eval_runtime": 434.6895,
|
| 1836 |
+
"eval_samples_per_second": 6.639,
|
| 1837 |
+
"eval_steps_per_second": 0.209,
|
| 1838 |
+
"learning_rate": 0.0001,
|
| 1839 |
+
"step": 29757
|
| 1840 |
+
},
|
| 1841 |
+
{
|
| 1842 |
+
"epoch": 109.89010989010988,
|
| 1843 |
+
"grad_norm": 0.3718918561935425,
|
| 1844 |
+
"learning_rate": 0.0001,
|
| 1845 |
+
"loss": 0.1479,
|
| 1846 |
+
"step": 30000
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"epoch": 110.0,
|
| 1850 |
+
"eval_accuracy": 0.2803187803187803,
|
| 1851 |
+
"eval_f1_macro": 0.6582487839550253,
|
| 1852 |
+
"eval_f1_micro": 0.7997274043785672,
|
| 1853 |
+
"eval_loss": 0.13255637884140015,
|
| 1854 |
+
"eval_roc_auc": 0.8626158546334878,
|
| 1855 |
+
"eval_runtime": 427.7015,
|
| 1856 |
+
"eval_samples_per_second": 6.748,
|
| 1857 |
+
"eval_steps_per_second": 0.213,
|
| 1858 |
+
"learning_rate": 0.0001,
|
| 1859 |
+
"step": 30030
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 111.0,
|
| 1863 |
+
"eval_accuracy": 0.2785862785862786,
|
| 1864 |
+
"eval_f1_macro": 0.6608614747058748,
|
| 1865 |
+
"eval_f1_micro": 0.8012935069355799,
|
| 1866 |
+
"eval_loss": 0.1319260448217392,
|
| 1867 |
+
"eval_roc_auc": 0.8637521073014844,
|
| 1868 |
+
"eval_runtime": 422.4227,
|
| 1869 |
+
"eval_samples_per_second": 6.832,
|
| 1870 |
+
"eval_steps_per_second": 0.215,
|
| 1871 |
+
"learning_rate": 0.0001,
|
| 1872 |
+
"step": 30303
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 111.72161172161172,
|
| 1876 |
+
"grad_norm": 0.3544025719165802,
|
| 1877 |
+
"learning_rate": 0.0001,
|
| 1878 |
+
"loss": 0.1466,
|
| 1879 |
+
"step": 30500
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 112.0,
|
| 1883 |
+
"eval_accuracy": 0.28101178101178104,
|
| 1884 |
+
"eval_f1_macro": 0.6595016342799644,
|
| 1885 |
+
"eval_f1_micro": 0.8019278738426415,
|
| 1886 |
+
"eval_loss": 0.13223350048065186,
|
| 1887 |
+
"eval_roc_auc": 0.8659084092462648,
|
| 1888 |
+
"eval_runtime": 420.8235,
|
| 1889 |
+
"eval_samples_per_second": 6.858,
|
| 1890 |
+
"eval_steps_per_second": 0.216,
|
| 1891 |
+
"learning_rate": 0.0001,
|
| 1892 |
+
"step": 30576
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 113.0,
|
| 1896 |
+
"eval_accuracy": 0.27997227997227997,
|
| 1897 |
+
"eval_f1_macro": 0.6592029124671744,
|
| 1898 |
+
"eval_f1_micro": 0.8024988392216453,
|
| 1899 |
+
"eval_loss": 0.13213913142681122,
|
| 1900 |
+
"eval_roc_auc": 0.8666766420318518,
|
| 1901 |
+
"eval_runtime": 423.8949,
|
| 1902 |
+
"eval_samples_per_second": 6.808,
|
| 1903 |
+
"eval_steps_per_second": 0.215,
|
| 1904 |
+
"learning_rate": 0.0001,
|
| 1905 |
+
"step": 30849
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 113.55311355311355,
|
| 1909 |
+
"grad_norm": 0.35069116950035095,
|
| 1910 |
+
"learning_rate": 0.0001,
|
| 1911 |
+
"loss": 0.1474,
|
| 1912 |
+
"step": 31000
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"epoch": 114.0,
|
| 1916 |
+
"eval_accuracy": 0.2823977823977824,
|
| 1917 |
+
"eval_f1_macro": 0.663088095209859,
|
| 1918 |
+
"eval_f1_micro": 0.8025030654094965,
|
| 1919 |
+
"eval_loss": 0.13204564154148102,
|
| 1920 |
+
"eval_roc_auc": 0.8661983610533127,
|
| 1921 |
+
"eval_runtime": 421.2287,
|
| 1922 |
+
"eval_samples_per_second": 6.851,
|
| 1923 |
+
"eval_steps_per_second": 0.216,
|
| 1924 |
+
"learning_rate": 0.0001,
|
| 1925 |
+
"step": 31122
|
| 1926 |
+
},
|
| 1927 |
+
{
|
| 1928 |
+
"epoch": 115.0,
|
| 1929 |
+
"eval_accuracy": 0.28378378378378377,
|
| 1930 |
+
"eval_f1_macro": 0.659797224924612,
|
| 1931 |
+
"eval_f1_micro": 0.8004266211604096,
|
| 1932 |
+
"eval_loss": 0.1319342404603958,
|
| 1933 |
+
"eval_roc_auc": 0.8625399730007867,
|
| 1934 |
+
"eval_runtime": 424.6871,
|
| 1935 |
+
"eval_samples_per_second": 6.796,
|
| 1936 |
+
"eval_steps_per_second": 0.214,
|
| 1937 |
+
"learning_rate": 0.0001,
|
| 1938 |
+
"step": 31395
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 115.38461538461539,
|
| 1942 |
+
"grad_norm": 0.29624369740486145,
|
| 1943 |
+
"learning_rate": 1e-05,
|
| 1944 |
+
"loss": 0.1468,
|
| 1945 |
+
"step": 31500
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 116.0,
|
| 1949 |
+
"eval_accuracy": 0.2844767844767845,
|
| 1950 |
+
"eval_f1_macro": 0.6627361818946377,
|
| 1951 |
+
"eval_f1_micro": 0.8022295974810655,
|
| 1952 |
+
"eval_loss": 0.13186337053775787,
|
| 1953 |
+
"eval_roc_auc": 0.8642598314802673,
|
| 1954 |
+
"eval_runtime": 423.8673,
|
| 1955 |
+
"eval_samples_per_second": 6.809,
|
| 1956 |
+
"eval_steps_per_second": 0.215,
|
| 1957 |
+
"learning_rate": 1e-05,
|
| 1958 |
+
"step": 31668
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"epoch": 117.0,
|
| 1962 |
+
"eval_accuracy": 0.28205128205128205,
|
| 1963 |
+
"eval_f1_macro": 0.6604165936303265,
|
| 1964 |
+
"eval_f1_micro": 0.8012607547491268,
|
| 1965 |
+
"eval_loss": 0.1317850947380066,
|
| 1966 |
+
"eval_roc_auc": 0.8634466760169507,
|
| 1967 |
+
"eval_runtime": 419.012,
|
| 1968 |
+
"eval_samples_per_second": 6.888,
|
| 1969 |
+
"eval_steps_per_second": 0.217,
|
| 1970 |
+
"learning_rate": 1e-05,
|
| 1971 |
+
"step": 31941
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 117.21611721611721,
|
| 1975 |
+
"grad_norm": 0.28633400797843933,
|
| 1976 |
+
"learning_rate": 1e-05,
|
| 1977 |
+
"loss": 0.1455,
|
| 1978 |
+
"step": 32000
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 118.0,
|
| 1982 |
+
"eval_accuracy": 0.2796257796257796,
|
| 1983 |
+
"eval_f1_macro": 0.6590147410119703,
|
| 1984 |
+
"eval_f1_micro": 0.8002395926924228,
|
| 1985 |
+
"eval_loss": 0.13159342110157013,
|
| 1986 |
+
"eval_roc_auc": 0.8616373075259771,
|
| 1987 |
+
"eval_runtime": 419.8006,
|
| 1988 |
+
"eval_samples_per_second": 6.875,
|
| 1989 |
+
"eval_steps_per_second": 0.217,
|
| 1990 |
+
"learning_rate": 1e-05,
|
| 1991 |
+
"step": 32214
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 119.0,
|
| 1995 |
+
"eval_accuracy": 0.28274428274428276,
|
| 1996 |
+
"eval_f1_macro": 0.6608406822787987,
|
| 1997 |
+
"eval_f1_micro": 0.8036745185622182,
|
| 1998 |
+
"eval_loss": 0.1319129317998886,
|
| 1999 |
+
"eval_roc_auc": 0.8678011174197509,
|
| 2000 |
+
"eval_runtime": 423.7674,
|
| 2001 |
+
"eval_samples_per_second": 6.81,
|
| 2002 |
+
"eval_steps_per_second": 0.215,
|
| 2003 |
+
"learning_rate": 1e-05,
|
| 2004 |
+
"step": 32487
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 119.04761904761905,
|
| 2008 |
+
"grad_norm": 0.31120315194129944,
|
| 2009 |
+
"learning_rate": 1e-05,
|
| 2010 |
+
"loss": 0.1451,
|
| 2011 |
+
"step": 32500
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 120.0,
|
| 2015 |
+
"eval_accuracy": 0.28135828135828134,
|
| 2016 |
+
"eval_f1_macro": 0.6614581971670047,
|
| 2017 |
+
"eval_f1_micro": 0.803593372600534,
|
| 2018 |
+
"eval_loss": 0.13164088129997253,
|
| 2019 |
+
"eval_roc_auc": 0.8661674020983411,
|
| 2020 |
+
"eval_runtime": 420.709,
|
| 2021 |
+
"eval_samples_per_second": 6.86,
|
| 2022 |
+
"eval_steps_per_second": 0.216,
|
| 2023 |
+
"learning_rate": 1e-05,
|
| 2024 |
+
"step": 32760
|
| 2025 |
+
},
|
| 2026 |
+
{
|
| 2027 |
+
"epoch": 120.87912087912088,
|
| 2028 |
+
"grad_norm": 0.31770700216293335,
|
| 2029 |
+
"learning_rate": 1e-05,
|
| 2030 |
+
"loss": 0.1454,
|
| 2031 |
+
"step": 33000
|
| 2032 |
+
},
|
| 2033 |
+
{
|
| 2034 |
+
"epoch": 121.0,
|
| 2035 |
+
"eval_accuracy": 0.28101178101178104,
|
| 2036 |
+
"eval_f1_macro": 0.6610641151618838,
|
| 2037 |
+
"eval_f1_micro": 0.8012604863092451,
|
| 2038 |
+
"eval_loss": 0.13184630870819092,
|
| 2039 |
+
"eval_roc_auc": 0.8635064611392681,
|
| 2040 |
+
"eval_runtime": 422.0264,
|
| 2041 |
+
"eval_samples_per_second": 6.838,
|
| 2042 |
+
"eval_steps_per_second": 0.216,
|
| 2043 |
+
"learning_rate": 1e-05,
|
| 2044 |
+
"step": 33033
|
| 2045 |
+
},
|
| 2046 |
+
{
|
| 2047 |
+
"epoch": 122.0,
|
| 2048 |
+
"eval_accuracy": 0.2817047817047817,
|
| 2049 |
+
"eval_f1_macro": 0.6647378818356079,
|
| 2050 |
+
"eval_f1_micro": 0.8049611099432415,
|
| 2051 |
+
"eval_loss": 0.13215216994285583,
|
| 2052 |
+
"eval_roc_auc": 0.8691576105910745,
|
| 2053 |
+
"eval_runtime": 436.9114,
|
| 2054 |
+
"eval_samples_per_second": 6.605,
|
| 2055 |
+
"eval_steps_per_second": 0.208,
|
| 2056 |
+
"learning_rate": 1e-05,
|
| 2057 |
+
"step": 33306
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 122.71062271062272,
|
| 2061 |
+
"grad_norm": 0.22290275990962982,
|
| 2062 |
+
"learning_rate": 1e-05,
|
| 2063 |
+
"loss": 0.145,
|
| 2064 |
+
"step": 33500
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 123.0,
|
| 2068 |
+
"eval_accuracy": 0.2817047817047817,
|
| 2069 |
+
"eval_f1_macro": 0.6604978306251739,
|
| 2070 |
+
"eval_f1_micro": 0.8010107932156931,
|
| 2071 |
+
"eval_loss": 0.13187836110591888,
|
| 2072 |
+
"eval_roc_auc": 0.8617537926061216,
|
| 2073 |
+
"eval_runtime": 431.3938,
|
| 2074 |
+
"eval_samples_per_second": 6.69,
|
| 2075 |
+
"eval_steps_per_second": 0.211,
|
| 2076 |
+
"learning_rate": 1e-05,
|
| 2077 |
+
"step": 33579
|
| 2078 |
+
},
|
| 2079 |
+
{
|
| 2080 |
+
"epoch": 124.0,
|
| 2081 |
+
"eval_accuracy": 0.2806652806652807,
|
| 2082 |
+
"eval_f1_macro": 0.6621515776947642,
|
| 2083 |
+
"eval_f1_micro": 0.8018739352640545,
|
| 2084 |
+
"eval_loss": 0.13141389191150665,
|
| 2085 |
+
"eval_roc_auc": 0.8638029186192627,
|
| 2086 |
+
"eval_runtime": 430.2675,
|
| 2087 |
+
"eval_samples_per_second": 6.707,
|
| 2088 |
+
"eval_steps_per_second": 0.211,
|
| 2089 |
+
"learning_rate": 1e-05,
|
| 2090 |
+
"step": 33852
|
| 2091 |
+
},
|
| 2092 |
+
{
|
| 2093 |
+
"epoch": 124.54212454212454,
|
| 2094 |
+
"grad_norm": 0.27631625533103943,
|
| 2095 |
+
"learning_rate": 1e-05,
|
| 2096 |
+
"loss": 0.1459,
|
| 2097 |
+
"step": 34000
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 125.0,
|
| 2101 |
+
"eval_accuracy": 0.2862092862092862,
|
| 2102 |
+
"eval_f1_macro": 0.6640721616133445,
|
| 2103 |
+
"eval_f1_micro": 0.804345987993574,
|
| 2104 |
+
"eval_loss": 0.13139639794826508,
|
| 2105 |
+
"eval_roc_auc": 0.8672404491355638,
|
| 2106 |
+
"eval_runtime": 432.0509,
|
| 2107 |
+
"eval_samples_per_second": 6.68,
|
| 2108 |
+
"eval_steps_per_second": 0.211,
|
| 2109 |
+
"learning_rate": 1e-05,
|
| 2110 |
+
"step": 34125
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 126.0,
|
| 2114 |
+
"eval_accuracy": 0.2862092862092862,
|
| 2115 |
+
"eval_f1_macro": 0.663003919720051,
|
| 2116 |
+
"eval_f1_micro": 0.804212663367593,
|
| 2117 |
+
"eval_loss": 0.13103623688220978,
|
| 2118 |
+
"eval_roc_auc": 0.8670350710768244,
|
| 2119 |
+
"eval_runtime": 432.4499,
|
| 2120 |
+
"eval_samples_per_second": 6.674,
|
| 2121 |
+
"eval_steps_per_second": 0.21,
|
| 2122 |
+
"learning_rate": 1e-05,
|
| 2123 |
+
"step": 34398
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 126.37362637362638,
|
| 2127 |
+
"grad_norm": 0.3177105188369751,
|
| 2128 |
+
"learning_rate": 1e-05,
|
| 2129 |
+
"loss": 0.1439,
|
| 2130 |
+
"step": 34500
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 127.0,
|
| 2134 |
+
"eval_accuracy": 0.28586278586278585,
|
| 2135 |
+
"eval_f1_macro": 0.6597731906072118,
|
| 2136 |
+
"eval_f1_micro": 0.8038346213944846,
|
| 2137 |
+
"eval_loss": 0.13152988255023956,
|
| 2138 |
+
"eval_roc_auc": 0.8672624342859965,
|
| 2139 |
+
"eval_runtime": 431.3827,
|
| 2140 |
+
"eval_samples_per_second": 6.69,
|
| 2141 |
+
"eval_steps_per_second": 0.211,
|
| 2142 |
+
"learning_rate": 1e-05,
|
| 2143 |
+
"step": 34671
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 128.0,
|
| 2147 |
+
"eval_accuracy": 0.2869022869022869,
|
| 2148 |
+
"eval_f1_macro": 0.668197478893632,
|
| 2149 |
+
"eval_f1_micro": 0.8042412977357216,
|
| 2150 |
+
"eval_loss": 0.13113313913345337,
|
| 2151 |
+
"eval_roc_auc": 0.8674002874836755,
|
| 2152 |
+
"eval_runtime": 439.4627,
|
| 2153 |
+
"eval_samples_per_second": 6.567,
|
| 2154 |
+
"eval_steps_per_second": 0.207,
|
| 2155 |
+
"learning_rate": 1e-05,
|
| 2156 |
+
"step": 34944
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 128.2051282051282,
|
| 2160 |
+
"grad_norm": 0.2520149350166321,
|
| 2161 |
+
"learning_rate": 1e-05,
|
| 2162 |
+
"loss": 0.1446,
|
| 2163 |
+
"step": 35000
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 129.0,
|
| 2167 |
+
"eval_accuracy": 0.28274428274428276,
|
| 2168 |
+
"eval_f1_macro": 0.6652814888251478,
|
| 2169 |
+
"eval_f1_micro": 0.8034694309287074,
|
| 2170 |
+
"eval_loss": 0.13096605241298676,
|
| 2171 |
+
"eval_roc_auc": 0.8665332355380903,
|
| 2172 |
+
"eval_runtime": 443.7844,
|
| 2173 |
+
"eval_samples_per_second": 6.503,
|
| 2174 |
+
"eval_steps_per_second": 0.205,
|
| 2175 |
+
"learning_rate": 1e-05,
|
| 2176 |
+
"step": 35217
|
| 2177 |
+
},
|
| 2178 |
+
{
|
| 2179 |
+
"epoch": 130.0,
|
| 2180 |
+
"eval_accuracy": 0.28655578655578656,
|
| 2181 |
+
"eval_f1_macro": 0.6657375892895663,
|
| 2182 |
+
"eval_f1_micro": 0.8034491503931017,
|
| 2183 |
+
"eval_loss": 0.1310083270072937,
|
| 2184 |
+
"eval_roc_auc": 0.866799015752045,
|
| 2185 |
+
"eval_runtime": 440.6588,
|
| 2186 |
+
"eval_samples_per_second": 6.549,
|
| 2187 |
+
"eval_steps_per_second": 0.207,
|
| 2188 |
+
"learning_rate": 1e-05,
|
| 2189 |
+
"step": 35490
|
| 2190 |
+
},
|
| 2191 |
+
{
|
| 2192 |
+
"epoch": 130.03663003663004,
|
| 2193 |
+
"grad_norm": 0.2916598916053772,
|
| 2194 |
+
"learning_rate": 1e-05,
|
| 2195 |
+
"loss": 0.1449,
|
| 2196 |
+
"step": 35500
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 131.0,
|
| 2200 |
+
"eval_accuracy": 0.2834372834372834,
|
| 2201 |
+
"eval_f1_macro": 0.6709132204127336,
|
| 2202 |
+
"eval_f1_micro": 0.8052362171687506,
|
| 2203 |
+
"eval_loss": 0.13133247196674347,
|
| 2204 |
+
"eval_roc_auc": 0.8699004377177725,
|
| 2205 |
+
"eval_runtime": 446.7612,
|
| 2206 |
+
"eval_samples_per_second": 6.46,
|
| 2207 |
+
"eval_steps_per_second": 0.204,
|
| 2208 |
+
"learning_rate": 1e-05,
|
| 2209 |
+
"step": 35763
|
| 2210 |
+
},
|
| 2211 |
+
{
|
| 2212 |
+
"epoch": 131.86813186813185,
|
| 2213 |
+
"grad_norm": 0.3473760783672333,
|
| 2214 |
+
"learning_rate": 1e-05,
|
| 2215 |
+
"loss": 0.1442,
|
| 2216 |
+
"step": 36000
|
| 2217 |
+
},
|
| 2218 |
+
{
|
| 2219 |
+
"epoch": 132.0,
|
| 2220 |
+
"eval_accuracy": 0.2806652806652807,
|
| 2221 |
+
"eval_f1_macro": 0.6557913726655867,
|
| 2222 |
+
"eval_f1_micro": 0.7985562048814026,
|
| 2223 |
+
"eval_loss": 0.13149647414684296,
|
| 2224 |
+
"eval_roc_auc": 0.8595249758820619,
|
| 2225 |
+
"eval_runtime": 447.0484,
|
| 2226 |
+
"eval_samples_per_second": 6.456,
|
| 2227 |
+
"eval_steps_per_second": 0.204,
|
| 2228 |
+
"learning_rate": 1e-05,
|
| 2229 |
+
"step": 36036
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 133.0,
|
| 2233 |
+
"eval_accuracy": 0.28794178794178793,
|
| 2234 |
+
"eval_f1_macro": 0.6689392948255155,
|
| 2235 |
+
"eval_f1_micro": 0.8051816958277256,
|
| 2236 |
+
"eval_loss": 0.1311328113079071,
|
| 2237 |
+
"eval_roc_auc": 0.8691700049040701,
|
| 2238 |
+
"eval_runtime": 444.1217,
|
| 2239 |
+
"eval_samples_per_second": 6.498,
|
| 2240 |
+
"eval_steps_per_second": 0.205,
|
| 2241 |
+
"learning_rate": 1e-05,
|
| 2242 |
+
"step": 36309
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 133.6996336996337,
|
| 2246 |
+
"grad_norm": 0.2959079444408417,
|
| 2247 |
+
"learning_rate": 1e-05,
|
| 2248 |
+
"loss": 0.1443,
|
| 2249 |
+
"step": 36500
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 134.0,
|
| 2253 |
+
"eval_accuracy": 0.28274428274428276,
|
| 2254 |
+
"eval_f1_macro": 0.6648386499372343,
|
| 2255 |
+
"eval_f1_micro": 0.802060714437774,
|
| 2256 |
+
"eval_loss": 0.1308571696281433,
|
| 2257 |
+
"eval_roc_auc": 0.8639881626262637,
|
| 2258 |
+
"eval_runtime": 444.917,
|
| 2259 |
+
"eval_samples_per_second": 6.487,
|
| 2260 |
+
"eval_steps_per_second": 0.205,
|
| 2261 |
+
"learning_rate": 1e-05,
|
| 2262 |
+
"step": 36582
|
| 2263 |
+
},
|
| 2264 |
+
{
|
| 2265 |
+
"epoch": 135.0,
|
| 2266 |
+
"eval_accuracy": 0.2869022869022869,
|
| 2267 |
+
"eval_f1_macro": 0.6684163123065296,
|
| 2268 |
+
"eval_f1_micro": 0.8038277511961722,
|
| 2269 |
+
"eval_loss": 0.13148072361946106,
|
| 2270 |
+
"eval_roc_auc": 0.8665118674205556,
|
| 2271 |
+
"eval_runtime": 437.5153,
|
| 2272 |
+
"eval_samples_per_second": 6.596,
|
| 2273 |
+
"eval_steps_per_second": 0.208,
|
| 2274 |
+
"learning_rate": 1e-05,
|
| 2275 |
+
"step": 36855
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 135.53113553113553,
|
| 2279 |
+
"grad_norm": 0.3723543882369995,
|
| 2280 |
+
"learning_rate": 1e-05,
|
| 2281 |
+
"loss": 0.1438,
|
| 2282 |
+
"step": 37000
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 136.0,
|
| 2286 |
+
"eval_accuracy": 0.28274428274428276,
|
| 2287 |
+
"eval_f1_macro": 0.659009971789042,
|
| 2288 |
+
"eval_f1_micro": 0.8024591213764248,
|
| 2289 |
+
"eval_loss": 0.13150115311145782,
|
| 2290 |
+
"eval_roc_auc": 0.8634352340808195,
|
| 2291 |
+
"eval_runtime": 444.5109,
|
| 2292 |
+
"eval_samples_per_second": 6.493,
|
| 2293 |
+
"eval_steps_per_second": 0.205,
|
| 2294 |
+
"learning_rate": 1e-05,
|
| 2295 |
+
"step": 37128
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 137.0,
|
| 2299 |
+
"eval_accuracy": 0.28586278586278585,
|
| 2300 |
+
"eval_f1_macro": 0.6666808903899752,
|
| 2301 |
+
"eval_f1_micro": 0.8035592643051771,
|
| 2302 |
+
"eval_loss": 0.1310679018497467,
|
| 2303 |
+
"eval_roc_auc": 0.8648124783367798,
|
| 2304 |
+
"eval_runtime": 434.2661,
|
| 2305 |
+
"eval_samples_per_second": 6.646,
|
| 2306 |
+
"eval_steps_per_second": 0.21,
|
| 2307 |
+
"learning_rate": 1e-05,
|
| 2308 |
+
"step": 37401
|
| 2309 |
+
},
|
| 2310 |
+
{
|
| 2311 |
+
"epoch": 137.36263736263737,
|
| 2312 |
+
"grad_norm": 0.36766815185546875,
|
| 2313 |
+
"learning_rate": 1e-05,
|
| 2314 |
+
"loss": 0.1452,
|
| 2315 |
+
"step": 37500
|
| 2316 |
+
},
|
| 2317 |
+
{
|
| 2318 |
+
"epoch": 138.0,
|
| 2319 |
+
"eval_accuracy": 0.2844767844767845,
|
| 2320 |
+
"eval_f1_macro": 0.6665598962110765,
|
| 2321 |
+
"eval_f1_micro": 0.8035426731078905,
|
| 2322 |
+
"eval_loss": 0.13124705851078033,
|
| 2323 |
+
"eval_roc_auc": 0.8661277510277622,
|
| 2324 |
+
"eval_runtime": 434.1413,
|
| 2325 |
+
"eval_samples_per_second": 6.648,
|
| 2326 |
+
"eval_steps_per_second": 0.21,
|
| 2327 |
+
"learning_rate": 1e-05,
|
| 2328 |
+
"step": 37674
|
| 2329 |
+
},
|
| 2330 |
+
{
|
| 2331 |
+
"epoch": 139.0,
|
| 2332 |
+
"eval_accuracy": 0.28967428967428965,
|
| 2333 |
+
"eval_f1_macro": 0.6661043989752415,
|
| 2334 |
+
"eval_f1_micro": 0.8052538519828238,
|
| 2335 |
+
"eval_loss": 0.13104070723056793,
|
| 2336 |
+
"eval_roc_auc": 0.8689438757606943,
|
| 2337 |
+
"eval_runtime": 433.2581,
|
| 2338 |
+
"eval_samples_per_second": 6.661,
|
| 2339 |
+
"eval_steps_per_second": 0.21,
|
| 2340 |
+
"learning_rate": 1e-05,
|
| 2341 |
+
"step": 37947
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 139.19413919413918,
|
| 2345 |
+
"grad_norm": 0.35373228788375854,
|
| 2346 |
+
"learning_rate": 1e-05,
|
| 2347 |
+
"loss": 0.144,
|
| 2348 |
+
"step": 38000
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 140.0,
|
| 2352 |
+
"eval_accuracy": 0.2834372834372834,
|
| 2353 |
+
"eval_f1_macro": 0.663466069531375,
|
| 2354 |
+
"eval_f1_micro": 0.8020416843896214,
|
| 2355 |
+
"eval_loss": 0.13169734179973602,
|
| 2356 |
+
"eval_roc_auc": 0.8642539428402185,
|
| 2357 |
+
"eval_runtime": 435.0147,
|
| 2358 |
+
"eval_samples_per_second": 6.634,
|
| 2359 |
+
"eval_steps_per_second": 0.209,
|
| 2360 |
+
"learning_rate": 1e-05,
|
| 2361 |
+
"step": 38220
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 141.0,
|
| 2365 |
+
"eval_accuracy": 0.2875952875952876,
|
| 2366 |
+
"eval_f1_macro": 0.6687691213000826,
|
| 2367 |
+
"eval_f1_micro": 0.8046521463311481,
|
| 2368 |
+
"eval_loss": 0.13089434802532196,
|
| 2369 |
+
"eval_roc_auc": 0.867299000192085,
|
| 2370 |
+
"eval_runtime": 429.8469,
|
| 2371 |
+
"eval_samples_per_second": 6.714,
|
| 2372 |
+
"eval_steps_per_second": 0.212,
|
| 2373 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 2374 |
+
"step": 38493
|
| 2375 |
+
},
|
| 2376 |
+
{
|
| 2377 |
+
"epoch": 141.02564102564102,
|
| 2378 |
+
"grad_norm": 0.2815115451812744,
|
| 2379 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 2380 |
+
"loss": 0.1445,
|
| 2381 |
+
"step": 38500
|
| 2382 |
+
},
|
| 2383 |
+
{
|
| 2384 |
+
"epoch": 142.0,
|
| 2385 |
+
"eval_accuracy": 0.28586278586278585,
|
| 2386 |
+
"eval_f1_macro": 0.6642894279153319,
|
| 2387 |
+
"eval_f1_micro": 0.8041640110473762,
|
| 2388 |
+
"eval_loss": 0.13103386759757996,
|
| 2389 |
+
"eval_roc_auc": 0.8657067870399482,
|
| 2390 |
+
"eval_runtime": 425.5573,
|
| 2391 |
+
"eval_samples_per_second": 6.782,
|
| 2392 |
+
"eval_steps_per_second": 0.214,
|
| 2393 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 2394 |
+
"step": 38766
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 142.85714285714286,
|
| 2398 |
+
"grad_norm": 0.3381010890007019,
|
| 2399 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 2400 |
+
"loss": 0.1441,
|
| 2401 |
+
"step": 39000
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 143.0,
|
| 2405 |
+
"eval_accuracy": 0.2872487872487873,
|
| 2406 |
+
"eval_f1_macro": 0.6623287859816251,
|
| 2407 |
+
"eval_f1_micro": 0.8019270122783083,
|
| 2408 |
+
"eval_loss": 0.13144278526306152,
|
| 2409 |
+
"eval_roc_auc": 0.8635436440782548,
|
| 2410 |
+
"eval_runtime": 433.7658,
|
| 2411 |
+
"eval_samples_per_second": 6.653,
|
| 2412 |
+
"eval_steps_per_second": 0.21,
|
| 2413 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 2414 |
+
"step": 39039
|
| 2415 |
+
},
|
| 2416 |
+
{
|
| 2417 |
+
"epoch": 144.0,
|
| 2418 |
+
"eval_accuracy": 0.28378378378378377,
|
| 2419 |
+
"eval_f1_macro": 0.6647534218687892,
|
| 2420 |
+
"eval_f1_micro": 0.8024974515800204,
|
| 2421 |
+
"eval_loss": 0.1311902105808258,
|
| 2422 |
+
"eval_roc_auc": 0.8649097280870156,
|
| 2423 |
+
"eval_runtime": 446.8955,
|
| 2424 |
+
"eval_samples_per_second": 6.458,
|
| 2425 |
+
"eval_steps_per_second": 0.204,
|
| 2426 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 2427 |
+
"step": 39312
|
| 2428 |
+
},
|
| 2429 |
+
{
|
| 2430 |
+
"epoch": 144.0,
|
| 2431 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 2432 |
+
"step": 39312,
|
| 2433 |
+
"total_flos": 1.3598709030716368e+20,
|
| 2434 |
+
"train_loss": 0.157796386979584,
|
| 2435 |
+
"train_runtime": 249885.5342,
|
| 2436 |
+
"train_samples_per_second": 5.232,
|
| 2437 |
+
"train_steps_per_second": 0.164
|
| 2438 |
+
}
|
| 2439 |
+
],
|
| 2440 |
+
"logging_steps": 500,
|
| 2441 |
+
"max_steps": 40950,
|
| 2442 |
+
"num_input_tokens_seen": 0,
|
| 2443 |
+
"num_train_epochs": 150,
|
| 2444 |
+
"save_steps": 500,
|
| 2445 |
+
"stateful_callbacks": {
|
| 2446 |
+
"EarlyStoppingCallback": {
|
| 2447 |
+
"args": {
|
| 2448 |
+
"early_stopping_patience": 10,
|
| 2449 |
+
"early_stopping_threshold": 0.0
|
| 2450 |
+
},
|
| 2451 |
+
"attributes": {
|
| 2452 |
+
"early_stopping_patience_counter": 0
|
| 2453 |
+
}
|
| 2454 |
+
},
|
| 2455 |
+
"TrainerControl": {
|
| 2456 |
+
"args": {
|
| 2457 |
+
"should_epoch_stop": false,
|
| 2458 |
+
"should_evaluate": false,
|
| 2459 |
+
"should_log": false,
|
| 2460 |
+
"should_save": true,
|
| 2461 |
+
"should_training_stop": true
|
| 2462 |
+
},
|
| 2463 |
+
"attributes": {}
|
| 2464 |
+
}
|
| 2465 |
+
},
|
| 2466 |
+
"total_flos": 1.3598709030716368e+20,
|
| 2467 |
+
"train_batch_size": 32,
|
| 2468 |
+
"trial_name": null,
|
| 2469 |
+
"trial_params": null
|
| 2470 |
+
}
|